PAPERS TO BE PRESENTED AT THE INTERNATIONAL CONFERENCE ON ATMOSPHERIC AND SPACE ELECTRICITY, MONTREUX, SWITZERLAND, 6-10 MAY 63

• Document Type: FOIA
• Collection: FOIA Collection
• Document Number (FOIA) /ESDN (CREST): 06461858
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 347
• Document Creation Date: December 28, 2022
• Document Release Date: September 25, 2017
• Case Number: F-2015-01649
• Publication Date: May 9, 1963

Approved for Release: 2017/09/11 C06461858
INFORMATION REPORT INFORMATION REPORT
CENTRAL INTELLIGENCE AGENCY
This material contains information affecting the National Defense of the United States within the meaning of the Espionoge Laws, Title
18, U.S.C. Sea. 793 and 794. the transmission Of revelation of which in any manner to an unauthorized person is prohibited by law.
COUNTRY
SUBJECT
DATE OF
INFO.
PLACE & -- 11/4) Lnr C3 prior
DATE ACQ.
Internptionalls;v A:6e,
REPORT NO.
(b)(3
aelers to 7...e Presented at the
DATE DISTR.
9 May 63
International Conference on
:.U4ospheric and Opace Electricity,
NO. PAGES 2
::ontreux, Switzerland, 6-10 !Iay 63
(b)(3
REFERENCES
19 _...7)r (,3 C: prior
THIS Is UNEVALUATED INFORMATION
5
4
3
2
1
'STATE
listed below are the titled and authors of the
foreign papers to be presented at the International Conference on � .
Atmosnheric and Space Electricity, Montreux, Switzerland, 6-10 Ney 1963:
(a) "Atmospheric Electricity Research in the Far East" by U.
Uatakeyama
(b) ''Report on Atmospheric Electricity in Central Europe 1959-62"
by R tiMaeisen
(c) "Atmosplleric alectricity Research in Great Britain, IrelanC.,
Africa and new Zealand" by WC A Nutchinson
(d) 'Electromagnetic Energy Padisted From Lightning" by Atsushi
Nianera
(e) "The Concepts of Atmos,heric Llectricity as Applied to the
Ionosphere" by IC flaeda
(f) "Ceoelektrische PrObleme der Blitzforschung" by Volker Fritsch
(g) "Charge Ceneration in Thunderstorms" by J Alan Chalmers
(h) "Relations Between Lightning Discharges and Different Types
of Musical Atmospherics" by Harald Norinder
(1) "Problems of Fair Weather Electricity; Introducing Remarks"
by 11 Israil
(j) "Action of Radioactivity and of Pollution upon Parameters of
Atmospheric Electricity" by J Bricard
CO "Generation of Electric Charges Outside Thunderclouds" by
Alan Chalners
4
3
2
1
Si., 1
MIMI from adantle
Inavallag ail
doclustication
1AP.MY
(NAVY f I AIR �
(FBI I IAEC
I I I I
INFORMATION REPORT INFORMATION REPORT
CONTROLLED
NO DISSE3t ABROAD
NO FOREIGN DISSSM
=SEM: The dissemination of this document is limited to civilian emplo�es and active duty military personnel/ within the intelligence components
of the EMIR member agencies, and to those senior officials of the member agencW who must act upon the information. However, unless specifically controlled
in accordance with paragraph B of DCID I/7, it may be released to those components of the departments and agencies of the U. S. Government directly
participating in the production of National Intelligence. IT SHALL NOT BE DlSSEMINATED TO CONTRACTORS. It shall not be disseminated to *Maniac-
tions or perm:mei, including consultants, under a contractual relationship to the U.S. Government without the written permission of the originator.
(b)(3
(b)(3)
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
(�.0.14- .D.E
- 2 -
(1) "Charge Generation in Thtmderstroms" by B J Mason
(m) 'the Theory of Lightning" by D J Malan
(u) 'types of Lightning" by N Kitagava
(o) "Lightning Protection" by D
(p) 'Whistlers as a Phenomenon to Study Space Electricity"
by N D Clarence
(b)(3)
UNCLASSIFIED.3
- end
(b)(3)
(b)(3)
pproved for Release: 2017/09/11 C06461858
SESSION
0- f Approved for Release: 2017/09/11 C06461856
Atmospheric Electricity Research in the Far East
H. Hatakeyama
. Atmospheric electric observations in the upper atmosphere.
Observations of the electrical conductivity and the electric field
in the upper atmosphere were made twice a day during the World Meteoro-
logical Intervals in the IGY and IGC at four stations in Japan--Sapporo
(43�03'N, 141�20,E), Tateno (36�031N, 140�08'E), Hachijojima (334071N,
139�471E) and Kagoshima (31�381N. 130�36'E).
The results of observations were discussed by K. Uchikawa (1961)
and ^showsthat the mean vertical distribution of the conductivity obtained
at respective stations are almost equal to each other from the ground up
to 500 nib level and that the conductivity increases with the latitude
in the upper troposphere and in the lower stratosphere as shown graphi-
cally in Fig. 1. This suggests that the conductivity in the free atmos-
phere is mainly under the control of cosmic ray intensity, because the
total intensity of cosmic rays increases with the ceomagnetic latitude.
Mean values of the potential gradient in the upper atmosphere
.
obtained at four stations are shown in Fig. 2. Values of the potential
gradient are large near the ground and decrease exponentially with an
altitude. Above 100 nib level the value becomes lower than 5 V/m. However,
as the accuracy of the measurement is 1: 5 V/m, the electric field
intensity above 1C0 mb level could not measured precisely.
When the exchange layer develops, as shown in Fig. 3, the sudden
decrease in the atmospheric electric field and relatively small sudden
pproved for Release: 2017/09/11 C06461858
teX.,
Approved for Release: 2017/09/11 C06461858
increase in the electrical conductivity were observed at the top of the
exchange layer as in other measurements in foreign countries. This
marked decrease and increase disappear for several days after the low
pressure or the cyclone passes through the observation point because the
exchange layer fades away after the low pressure.
The conduction current in the exchange layer is about 1.3 times
lapger than that above the exchange layer; This means that the "Austauschn
contributes to generate the conduction current in the exchange layer
through the rapid production of ions.
Day to day variations of the electric field and the electrical
conductivity in the upper troposphere were investigated. These varia-
tions are related to meteorological factors, such as the air mass, the
upward or downward motion of the air, and the jet stream. For example
the increasing rate of the electrical conductivity with respect to the
latitude in case of a strong jet stream is larger than that ir a weak jet.
The concentration of small ions was computed from the observed elec-
trical conductivity. The vertical air currents in and around the jet
stream were calculated using the time and space variation of the concen-
tration of mall ions. When the intensity of the jet stream is increasing,
downward motions of the air predominate in and around the jet stream,
only except the southern and lower part of it where upward motion exists.
And vice versa when the intensity of the jet stream is decreasing. An
example of which is shown in Fig. 4. These characteristics coincide
qualitatively with those obtained thermodynamically from the time and
- 2 -
pproved for Release: 2017/09/11 C06461858
;!.
Approved for Release: 2017/09/11 C06461858
space variations of the air temperature.
At Poona in India, observations of the potential gradient using
s
radiovmde techniques have been made since 1953 and systematic observa-
tions were made during the IGY. The results of the observations were
discussed by S. P. Venkiteshwaran and Anna Mani (1962).
During clear weather, in both winter and summer, the higher values
of potential gradient are confined to a region extending from the ground
to about 600-500 nib, above which height it either remains fairly constant,
at about 20 V/m, or increase slightly with height. Within the exchange
layer, there are appreciable diurnal variations in the potential gradient.
They are at a minimum and almost constant during the hotter parts of the
day and higher at other times.
The data obtained during the IGY 1957-59 have been classified into
four groups, corresponding to the four main seasons -- (1) November-
February (Uinter), (2) March-May (Summer), (3) June-August (Monsoon
season) and (4) September-October (when the monsoon is withdrawing and
the skies are clear or partly covered with low clouds with a few thunder-
storms).
Examples of the results of observations are shown in Fig. 5.
In winter, high potential values are confined to a region from the
ground up to about 600 mb (4.4 km), above which, which represents,,the
top of the austausch region, the potential gradient remains almost steady
(about 10-40 V/m) with increasing height up to about 300 nib (9.7km). Above
which it again increases up to the highest levels studied, -- 50 nib (19 km),
- 3 -
;
7 ti
pproved for Release: 2017/09/11 C06461858
r���
'KAY
Approved for Release: 2017/09/11 C06461858
suggesting an inci.eas in the particle content of the atmosphere. The
300 mb region nearly corresponds to the altitude of the jet stream over
,India to the north of Poona. The increase in the potential gradient
above 300 mb therefore suggests the existence of fine suspended particles,
presumably of extra terrestrial origin, in a larger concentration just
above the level where the extra-tropical stratosphere flows into the
troposphere, through the region of the jet stream between the tropical
and extra-tropical stratospheres.
The conditions over Poona in the summer season are as follows:
(a) The austausch region extends up to 500 mb (5.8 km), about 1.4 km higher
than in winter: (b) the region of maximum potential gradient lies very
near the ground and the potential gradient generally decreases with height
up to 500 mb: (c) above the austausch region, the potential gradient it
quite steady at the 500 mb value up to 150 mb (14.2 km) or above: and (d)
the potential gradient above 500 mb is of the order of 20-P0 V/m compared
to 10-40 V/M during winter.
The observations during the monsoon month are again markedly differ-
ent from those observed in either winter or summer. The potential gradient
attains its highest values between 800 and 600 mb, which is the region of
maximum Cloudiness during the season. The upper limit of the region of
high potential gradient (500 mb) also represents roughly the top of the
clouds during this season, The most important difference is the steady
fall in the potential gradient above 500 mb, from 20-60 Vim to about 20-
40 Vim at 200 mb, and decreasing to less than 10 V/m at About 100 mb.
- 4 -
'
ft!, �
:
i't-�4c�:"1:411::.
---..
7
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Tit
September and October is a transition period when the monsoon is
withdrawing and winter conditions are setting in. The potential gradient
values are characteristic of both seasons.
T. Sekigawa (1960) observed and discussed the atmospheric electric
potential gradient at the summit of Mt. Fuji (3,776 m). Results are shown
graphically in Fig. 6. In summer months (May-August), the potential
gradient is large in later afternoon hours and small at about midnight.
On the contrary, in winter months (November-February) it is large in early
a
morning hours and small in later morning hours. In equinoxial months
(March, April, September and October) the diurnal variation is double
oscillation and maxima appear after midnight and in later afternoon hours
and minima in the morning and in the evening hours.
The diurnal variation in winter corresponds to the universal change
of potential gradient (9h L. M. T. ----- Oh U. T.). In summer months the
top of the exchange layer exceeds the summit of Mt. Fuji, and the potential
gradient is higher in afternoon hours and smaller in night hours. And in
equinoicial months characteristic of both summer and winter appears.
2.. Atmospheric electric elements near the ground.
Kondo (1959) discussed the secular change of atmospheric electric
elements using the observational data at the Kakioka Magnetic Observatory
for the period 1930-1957. He found that the potential gradient is decreas-
ing and the electrical conductivity is increasing since 1953. He thought
the cause of this decrease in the potential gradient and increase in the
- 5 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
electrical conductivity was the artificial radioactivity of fallouts due
to the test explosions of nuclear bombs.
Secular variations of the potential gradient at Kakioka (36.141N,
140�111E) and Memambetsu (43Q551N, 144�121E) are shown in Fig. 7. The
curve of Kakioka is the deviation from the mean value 130 V/m for 1936-
49, and that if tor Memambetsu is that from the mean value 124 V/m for
1950-53. In the fall of 1958 the test explosions were stopped and the
potential gradient gradually recovered its normal value, but in the summer
of 1961 the test explosions were again started and the potential gradient
is decreasing speedily.
Hatakeyama and Kawano (1953) reported the diurnal change of the
potential gradient at several places in Japan. In Tokyo we have observa-
tions of that in rather old time 1897-1903, which is shown in Fig. 8. We
are making the observation of potential gradient in Tokyo since January,
1962. The mean diurnal variation January-August, 1962, is Shown also in
Fig. 8. This type of diurnal variation is usually seen in large cities.
In the upper part of Fig. 8, the mean diurnal variation at Kakioka
for the period 1936-55 is shown. The distance between Tokyo and Kakioka
is about 70 km and the type of the diurnal variation at Kakioka has never
changed up to the present. Sixty years ago, the air at Tokyo was very
clear and the type of the diurnal variation was rural.
Misaki (1950, 1961) devised a mothod for measuring the ion spectrum
and discussed the relation between the ion spectrum and the electrical
conductivity. According to his method the inner cylinder of aspiration
- 6 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-
t )
is observed
_ ) as ordinate
and potential applied to the outer cylinder as abscissa ( tA.ICz
are current and capacity of each part respectively). Ion spectrum is
obtained by deriving the first derivative of this characteristic curve and
the second derivative is not needed.
He made experiments for obtaining the mobility spectrum of atmospheric
ions in the mobility region between 3.0 and 0.2 ce/V.sec. in 1960 in the
polluted air at Tokyo and in the clear air at Karuizawa. Results of the
diurnal series of observations made at both sites indicate some effects of
pollution on the relation between the electrical conductivity and the
mobility spectrum. In the polluted air, scores of per cent of the conduc-
tivity is attributed to the large or the intermediate ions while the
conductivity in the clear air is practically attributed to the small ions
only, as is generally believed.
The spectrum of the small ions does not shift on the mobility axis,
maximum concentration lying in the interval 1.0-0.7 cmiVV.sec., regardless
of the variations in the conductivity. On the contrary, the equivalent
mobility, i. e. the ratio of the polar conductivity to the small ion con-
centration, changes with the variations of the conductivity in the intensely
polluted air. An example of the results of observations are shown in Fig.
2(
apparatus is divided into two parts. The value (r
and the characteristic curve is formed taking ( 411(ti
�1
9.
The ionizations by 0/-, (3-, T- rays and (3-, rays and the
- 7 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
natural radioactive dust concentration in the atmosphere near the ground
have been observed continuously with two ionization chambers and an electro-
static precipitator at Tanashi near Tokyo since April, 1958, by M. Kawano
and S. Nakatani (1958, 1959). They discussed the results of observations.
On fine days the diurnal variation of the ionization by 04--, (3-,
I"- rays is similar to that of the ionization by 3-, 1^- rays. As
is shown in Fig. 10, the maximum value occurs in early morning (4-6 h),
and the minimum in the daytime (11-13 h). On cloudy and rainy days the
time variations are very irregular and the values are considerably larger
than those on fine days. On fine days the values of ((3, t)/( 01, (3,1r)
are about 2-5 per cent, being large in the daytime and small at night, but
the values on cloudy and rainy days are considerably smaller than these
on fine days.
The natural radioactive dust concentration is large at night and
small in the daytime, and the diurnal variation being similar to that of
ionization. But the amplitude of the diurnal variation curve of the dust
concentration collected with the electrostatic precipitator is remarkably
larger than that of the ionization by f-s - rays measured with an
ionization chamber.
The results of simultaneous observations mentioned above seem to be
important for the researches on the natural radioactivity and on the fre-
quency distribution of the particle size of the radioactive dusts in the
atmosphere.
M. Kawano and S. Nakatani (1961) studied the size distribution of
dust particles suspended in tht atmosphere near the ground which carry the
-8-
:'
�
r �
I I �
pproved for Release: 2017/09/11 C06461858
;',
Approved for Release: 2017/09/11 C06461858
I
naturally occurring radioactive substances by the cascade impactor and
autoradiography. The cascade impactor was used for classifying the dust
particles into four groups by their particle sizes, and the autoradiography
was used for counting the number of �--tracks of each class at 0 hrs and
18 hrs after collection by the impactor.
Table. Size distribution of dust particles
which hold the 0(-activity.
Class
Particle size (r)
Number of r/ -tracks
after collection (per
Number of D(-tracks
after collection (per
at 0 hrs
unit area)
at 18 hrs
unit area)
2.5-0.9 0.9-0.5
10
0
-4
13
4
IV
0.5-0.3
200
17
According to the results of measurements, as shown in the Table, a
large part of naturally occurring radioactive dust was concentrated in the
size range below 0.5tA, and the radioactivity Was radiated almost solely
from radon and thoron daughter products of short half lives.
M. Kawane (1957) pointedf,the abnormal increase of the ionizatien by
AS is shown
obscuration
- rays was found during the solar eclipse on April 19, 1958.
in Fig. 11, the maximum value occurred at the time of maximum
and was mere than twice that on the ether days.
3: Atmospheric electric elements during disturbed weather.
Ch. Magone and K. Orikasa (1960, 1961) and K. Orikasa (1962) made
simultaneous observations of the surface electric fields, the charge on
- 9 -
4
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
raindrops and snow particles, the form of snow particles and the intensity
of rainfall and snowfall from 1956 to 1960 at Sapporo. And the latter
author made similar observations simultaneously at two stations 1.2 km
apart each other.
Analysing the data of these records, the following conclusions were
obtained. When the rainfall was light or steady, positive field relatively
smaller than those of fine weather or negative field was observed, and
when there was light or steady snowfall, positive field was observed. During
continuous heavy rain or snowfall and during heavy rain shower or snow
shower, wave patterns of the electric field were often observed.
In almost all the cases of positive or negative electric field and
t:t4.
of the wave patterns of field, mirror image relations 11,14 generally
between the sign of electric field and the sign of electric charge on rain
or snow particles. But in the beginning of rain or snowfall and when the
rapid increase in the intensity of rain or snowfall occurred, the sign of
the electric field and the sign of electric charge on particles because
the same. An example of the observation is shown in Fig. 12.
To explain the mirror image relation mentioned above, the author
considered that the rain or snow particles were mainly electrified in the
cloud and caled electric charges down to the ground, consequently the
cloud may be electrified to the sign opposite to the net charge which was
carried down to the ground by the particles. The case of the same sign of
both the electric field and the electric charge on rain or snow particles
was explained hypothetically by considering the space charge due to charged
- 10 -
Lt.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
rain or snow particles.
Kikuchi and Magono (1961, 1961a) measured charges on natural snow
crystals before and after their artificial melting during snowfall. It
was found that the snow crystals obtained considerable positive charge
when they were melted. This observation appears to explain the above
mentioned reneral observational fact that in steady rainfall negative
surface electric field is predominant and raindrops carry positive charge
in most cases, while in steady snowfall positive surface electric field
is predominant and most of the snow crystals are charged negatively.
T. Ogawa (1960) and T. Cgawa and S. Saga (1961) made the continuous
observations of the electric current carried by rain drops, the rate of
gr.�
!
4!:
J
rainfall and the surface potential gradient. froviding the Wilson's
_
-.. ,
� , A.
theory of ion capture by water drops,the raindrop starts a cloud with )
.1,
:IroP:
V
a small charge in the same sign ;43 the electric field and reverses its
sign at a point between the cloud base and the ground. A quantitative
representation between the rain current, the rate of rainfall and the
potential graclient was :Assumed and a relation between the surface poten-
tial gradient and the potential gradient in the charging region of rain-
drops below the clouds was deduced. An example of the effect of splashing
of raindrops at the ground was shown, in which the intensity of rainfall
was 10 mm/hr or more the effect reduced the surface potential gradient
in the value of about 2 V/cm toward negative.
Department of physics in the University of Singapore has Flans to
expand itself and a new Atmospherics Physics Laboratory will soon be
,P
� pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
built. Last Ldqd up to August of this year, under the supervision
of Hon Yung Sen, J. Fa]iam investigated the electrical conductivity of
the atrrosphere near the ground durrlt the disturbed weather and some
irteresting results were obtained but they are not yet published.
References
Hatakeyama, H. and ;:awano (1953): On the Diurnal Variation of Atmos-
pheric rote-ncial Gradient in the Japan Archipelago, rap. In Met.
and fleophys. 4, 55-60.
Kawano, M. (1957): The Result of Observation of the Rate of Ion Pair
rrod-oction in the Atmosphere during the Solar Eclipse, Apr. 19,
195, ourn. Geom. and Geoelec. 9, 210-211.
Kawano, M. and 3. Nakatani (1958): The Results of Routine Observations
of the Ionization and the Natural Radioactive Dust Concentration
in the Atmosphere in Tokyo, Journ. Net. Soc. Japan, 36, 135-140.
Kawano, M. and 3. Nakatani (1959):
The Absolute Measurement of the
Concentrations of the hadioaotive Substances in the Atmosphere in
Tokyo, Journ. Gemag. and Geoelec. 10, 56-63.
Kawano, M. and S. Nakatani (1961): Size 9istribution of Naturally Occurr-
ing Radioactive rust measured by a Cascade Impactor and Autoradio-
graphy, Geofis. rurti. n Apr'. 50, 243-248.
Kikuchi, K. and Ch. Maoono (1961): On the Electrification of Snow
Crystals by their Melting, Journ. Japanese Society of Snow and
Ice, 23, 41-45.
- 12 -
"
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Kikuchi, K. and C. Magono (1961a): On the Electrification of Snow Crys-
tals by their Melting, II, Journ. Japanese Society of Snow and Ice,
23, 155-158.
Kondo, G. (1959): The Recent Status of Secular Variations of the Atmos-
pheric Electric Elements and their Relation to the Nuclear Explosions,
Memoirs of Kakioka Magnetic Observatory, 9, 2-6.
Magono, Ch. and K. Orikasa (1960): On the Surface Electric Field During
Rainfall, Journ. Met. Soc. Japan, 38, 182-194.
Magono, Ch. and K. Orikasa (1961): On the Surface Electric Field Caused
by the Space Charge of Charged Raindrops, Journ. Met. Soc. Japan,
39, 1-11.
Misaki, M. (1950): A Method of Measuring the Ion Spectrum, Pap. in
Met. and Geophys. 1, 313-318.
Misaki, M. (1961): Studies on the Atmospheric Itia Spectrum (1 and 2),
Pap. in Met. and Geophys. 12, 247-276.
Ogawa, T. (1960): Electricity in Rain, Journ. Geomag. and Geoelec.
12, 21-31.
Ogawa, T. and S. Saga (1961): Charge on Rain Drop, Journ. Geomag. and
Geoelec. 12, 99.
Orikasa, K. (1962): On the Disturbance of the Surface Electric Field
Caused by Rainfall and Snowfall, Geophys. Bull. Hokkaido Univ.
9, 123-160.
Sekigawa, T. (1960): Observation of Atmospheric Electricity at the Summit
-13-
proved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
of Mt. Fuji, Tenki, 7, 65-71 (in Japanese).
Uehikawa, K. (1961): Atmospheric Electric Phenomena in the Upper Air
over Japan (Part I and II), Geophys. Mag. 30, 617-672.
Venkiteshwaran, S. P. and Anna Mani (19("?): Measurement of Electrical
Potential Gradient in the Free Atmosphere over Poona, Journ. Atm.
Sci. 19, 226-231.
-- �
pproved for Release: 26-17/0-671TC06461858
CA,
x^Zr, �
�'.:
� 2'0'
IOU
1000
Approved for Release: 2017/09/11 C06461858
Stowe
Tau=
Ha[11.1011012
-
10D 150 � 10-" d
Fig. 1. Mean values of the negative
polar conductivity.
pproved for Release: 2017/09/11 C06461858
1.i
�i�
� Approved for Release: 2017/09/11 C06461858
so
Ss
1.
Ha
toO
M
Fig. 2. Mean values of the potential
gradient. Sa: Sapporo,
Ta: Tateno, Ha: Hachijojima
Ya: Kagoshima
;.
:
� .1.4.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
1111
700
900
10D0
;14
-am'
sic
amildt/
Fig. 3. Ascent curves of the negative
polar conductivity ( ), The
potential gradient (E), the
temperature (T) and the
relative humidity (R,H,)
observed at Tateno.
- Approved for Release: 2017/09/11 C06461858
A
n�N
Approved for Release: 2017/09/11 C06461858
1.9
A
10 �a
7t
I /3.1 24
----- �
\
\
0:1 1.4
1.9
0;11
2.4
3.9
0.4 2.7� "
1;9
0.9 �
�.ter.
7.9
;
�
\ 6!s
;
�.
3r311 3014
,4111.1.511q,
�1,.
. - ---------
1E7
Fig. 4. Vertical air currents computed
from the time and space variation
of the concentration of small
ions. Arrow: The direction of
vertical motion (unit: cm/sec.).
Broken line: Isotach (knots).
Thick Line: tropopause.
In(
.d
4
4
A
`1. 4
i�j14
cPel
�� - - � � - c5.114 �A
A pproved for Release. 2017/09/11 CO6461858
Approved for Release: 2017/09/11 C06461858
A WINTER
enb
100 �
500
900
2141. Nov. 153 2024.0ot. 2:53 2025. Jon.22154 2034.Jon.23:54 1705 Nov 2257
CLEAR SKY CIA AR SKY CLEAR SKY CLEAR SKY Sc 2/8
B SUMMER
100 �
500
400
204. moc 4164 2050.144c 9!54 0812.iior. I054 2056.0.1or.104 0617.Mor. W54
Ac 3/10.C16 tr. CLEAR SAY At Ir. CLEAR SKY HAZY SKY
100
500
900
too
C MONSOON
1654. July. 16513 1648. July. 17:58 1614.tiog. 7.36 1720. Juni! 1659 1624. July 15..59
Fs. OVERCAST Cu4te.C1 t,. Cu 5/13.Ci it. Cu2/8 RA11%400-420)
POST MONSOON
TTm
2025.0cl. 21'.55 2014 Ott 235) 2016. Ot I. 28755 2025 Oct 3055 2025.0c, 51753
C�.C41411 C.2i0 Sczni.44441. Ci Ii. CLEAR SAY
Fig.5. Seasonal variation of potential gradient with
height at Poona. The horizontal scale of each
graph is logarithmic, with potential gradient
by radiosonde from 1 to 1,000 V/M.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Fig. 6. Diurnal variation of the potential gradient at
the summit of Mt. Fuji.
�i.141,-.1.4� l't� t::! A 11;:�����! �.fu.:iti..I5;
pproved for Release: 2017/09/11 C06461858
C
+40
+20
0
-20
-40
-60-
-80-
+40
+20
0 1950
-20-
-40
-60
Fig. 7. Secular variation of the potential gradient at
Kakioka and Memanbetsu.
Approved for Release: 2017/09/11 C06461858
Fig. 8. Diurnal variation of the potential gradient
at Kakioka and Tokyo.
regrt*,
"1, '4�1
.... .. 7 .. :F,Ir:77' � 16.4r
pproved for Release: 2017/09/11 C06461858
i i
4
IF 1 !I
1.11.1 I
11,F
II
1
TU.
I
II
,
I
I
....7 .. ,...,_
14L1
0:i III i
.1:1 I I : I I
Hi I I;
! ii
II
1 i
I
:I
11
1
""
--",h`
i.i4-1-:
Hi I
lii � 1 '
IJ
.
ill. a II
. II i!
Ili!
t'
1
---'."�i-
I I
!
I
dill.
.4
:.
i ;
!
j
!r-

IIIILIIilh_
. _
1

'
-F-fl'
:t
1 1 i -fl
1....
in
_-
1
in
1 i ..-
I t-r
i _
0
2 I 05 02 5 2 i" 05 02 5 2 1 05 02 5 2 1 05 02
NOBILITY Che/ VOLT. SEC
Fig. 9. Diurnal series of the Positive ion Spectra (Tokyo)
E;
<
CD
0_
o
7:1
CD
CD
CD
cD
n).
cD
Co
0
cD
co
CP
c9
13: Ono "
Fig. 10. The Diurnal Variation Curves of Ionization in the
-
Atmosphere .on Several fine Days. �
24k.
LMT
rate of
ion Pair production
mean values of
ordinary calm days.,
1-fh
L. rt. T.
Fig. 11. The Record of the Ion Pair Production During the Solar
Eclipse, April 19, 1958.
An Example of the Simultaneous Observation of the
Surface Electric Field, the Charge on Snow Particles
The Form ofSnow Particles and the Intensity of Snowfall
During Snow Showers at Two Stations 1.2 kin apart of each
other.
pproved for Release: 2017/09/11 006461858
" Approved for Release: 2017/09/11 C06461858
Report on Atmospheric Electricity in
The following summary is based on publications,
dispatched reports from "Atmospheric Electri-
ciens" in Central Europe. The author would re-
gret it deeply, if any publication would have
been disregarded, because it was not known to
him. On the other side the dispatched reports
and publications have been so numerous, that not
every one has been mentioned. The gravity has
been put on the new knowledge..
Disposition: The report is dividedibbovarious subjects:
General matters of atmospheric electricity,
phenomena in fine weather.
Atmospheric electric aerology, atmospheric
electric circuit and potential of the ionosphere.
Conductivity, ions, radioactivity.
Precipitation electricity.
Thunderstorms, lightnings, 'aeries., whistlers.
Electrical phenomena in space.
Biological action of atmospheric electricity.
1) General natters of atmospheric electricity phenomena
In the International Geophysical Year registrations of
the atmospheric electricity have been made dt numerous
stations such as: Arose, Payerne (Switzerland), Swidrze
(Poland), Murchischon-Bay (Spitzbergen), Hohenpeissen-
berg, Garmisch, Zugspitze, Black Forest, Eifel, Potsdam
(Germany) and others. The results of these measurements
�'A pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
- 2 -
are manyfold. The difficulties in separating the global and local
influence have been explained. Therefore it must be concluded that
atmospheric electrical registrations on the ground can give infor-
mations especially about the local events of atmospheric electricity
in some cases, but only at excellent places and very seldom about
the global events, such as the voltage between ionosphere and earth
and therewith the world thunderstorm activity or about a spacious
quality of the airmasses, such as the columnar resistance. The local
and the global part are about equally large. There are always some
informations not allowing an exact separation. This holds good for
the single case, but also very often for the temporal mean value.
Al,. the registration of atmospheric electricity by Saxer L.
and Sigrist 1) in Arose (1800 - 2650mabove NN) proves
that. Although Arose: is a place with great air-cleantess, the local
influence can be found. Only the station in 2650 m NN shows equa-
lised courses of the day in a few cases. Of some interest seems to
be the discovered course in parallel of field at the ground and
content of Oxon in the layer 0 - 20 km above the ground. The
explanation of the relation between the 03 and the columnar-resistance
is the following: The percentage of 03 increases as well as the
electric field in sink processes because the columnar-resistance
is getting less and the vertical-current-density is getting more.
Israel H. 2) brings the spacious atmospheric electric
phenomena in connection with geophysical effects, as radioactivity
of the atmosphere and the exchange. The atmospherk electric
fluctuation means a quality to him which can be exploited on the
synoptical way and can be used as an indication of the type of
exchange.
k�hleisen R. 3) has continued the investigations about
the atmospheric electric fluctuation at the coast of the German
lea. The strong and short periodical variations of field, vertical-
current and electric space-charge density have already to arise
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
� 3 �
on the open sea and will be caused by strong exchanges as it could be
proved by registration ". of temperature, water-vapour and the speed
of the wind and by observations of the low Sc-clouds at the same time.
The exchange ham to follow in big packages whereby the air near the
water must to be exchanged by fresh air from a height of 100 - 300
in a period of several minutes. The air near the water probably gets
a positive space-charge by the electrode-effect, which the air from
above does notpossees, and so the steep changes of the atmospheric
electric quality will arise at the fixed station.
4)
The electrode-effect has also been discovered over the lake of
Constance. The electric field E has been measured in various
altitudes from 0 - 100 m at fixed stations and with captive-
balloons. The value in an altitude of 10 m or more are just half
of the value above the water surface. The measured and the calculated
apace-charge-density is in agreement. If these results will be
transferred to the circumstances at the coast of the German sea, it
can be found that the peaks of the fieldstrength and the apace- .
charge-density are in accordance with the very constant values of
the lake of Constance.
Ninet C. 5) calculates a formula under consideration of
the "eddy diffusion", which describes the correlation of the
electric parameter of the atmosphere: space-charge, field-strength
and conductivity. By the assumption of Whipple about the convection
current this formula can be used in order to win the eddy diffusion
coefficient by the values of the electrical quantities measured
near the ground. The author has made these measurements and has
got the eddy diffusion coefficient as function of the temperature-
gradient, the speed of the wind and the Richardson number.
that
Ninet believes the values received by this method have come into
accordance with the experiences, although he made some confined
suppositions. This work is an valuable contribution to the problem
"electrode-effect".
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-4
Rei� ter R 6) . gives a summary about his registrations at the
Zugspitz-massif in the last 6 years. Besides others he tries to
explain some interesting observations. Be found out a relation between
the ratio of the small ions concentration n+/n_ and the increasing
speed of the wind. The herewith connected positive space-charge shall
!Lave the action of an exchange generator. In the evening decrease of
the wind velocity the positive space-charge also disappears and
herewith the atmospheric electric potential gradient sinks. That is
Reiter's explanation for the sunset effect.
Kilinski E. 7) disposedadifficulty with atmospheric electric
measurements. The impairing of the isolation by spiders' webs can
ed
be prevent by rotation of the antenna such as a round vertical current
plate.
2) Atmospheric electric aeroloRy, atmospheric electric circuit, and
potential of the ionosphere.
'gy
First there should be mentioned some of the general publications
8)
about the subject on atmospheric electric aerology.Israel R.
describes measuring methods and measuring results in the free atmosphere
in details. Be puts them in connection with the conditions near the
ground. and the meteorological conditions such as exchange, inver-
sions etc.
9)
A chapter by Georgii W. is devoted to the special atmo-
spheric electric measuring opportunity in a glider. In his institute
Reinhardt M. has instrumentated a doubleseat glider for meteorological
and atmospheric electric measurements. He is able to measure the
vertical component of the atmospheric electricity field with a field
mill and two radioactive collectors on the wing and under the wing.
Besides this the horizontal component of the potential gradient and
the positive and negative conductivity of the air can be measured
continuously during the flight. It can be reported about the
already existing results after a concentrated exploitation.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
5
Milhleisen R. and Fischer H.J. 10)report on
the difficulties in exact measuring the atmospheric electrical field
in the free atmosphere and its remedy, based on their numerous
investigations with balloons and captive balloons. A special starting
method has been developed in order to avoid a positive
the balloon team with triboelectrical effects at the
wire. A negative charging of the wire, what happened
charging of
suspending
always during
flights through the ice satisfied areas, has been cleared up by
captive-balloon investigations as charge formations at rime.
11) The captive-balloon measurements by Lugeon J. give
further explanations about the relation of the field and the
conductivity in the lower atmosphere. The electrical quantities
make possible an exact determination of the limits of atmospheric
mist layers; they can be determined up to about 10 m by the
ductivity. This is more exact than it is possible to derive
the course of temperature and humidity. It has been noticed
con-
from
that
the meteorological quantities show these mist layers scarcely
or even not at all.
An exploitation of the ,Atmospheric electric work made by
Lugeon J., Junod A., Wasserfallen P.
and Rieker J. 12)in the IGY gives some new and precise
material about the course of the field and conductivity in the
free atmosphere. These mean values given in the following table
have been worked out f1m228 atmospheric electric and 33 conductivity
measurements above the Murchison-Bay and 1'0095 atmospheric electric
and 81 conductivity sondages above Payerne (see table 1).
A comparison with former, values also from other authors demonstrates
that the field values
it had been assumpted
worked out by M ii h
in the stratosphere are eminently lower than
in earlier times. These facts have also been
leisen R. and Fischer H.J.13)
Their mean values became less as well shown in table 1.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
h (km)
NN
dVidm
Murch.B.
(V/m) 1)
Payerne
1,- f(10-142
Murch.B.
i
-12-1) 1)
Payerne
. ,
2)
dVds(Via)
Weissenan
0,007
89,0
-
1,70
-
-
0,49
-
107,3
-
2,55
-
1,0
53,1
98,3
3,27
3,07
104
3
24,5
50,7
5,13
5,32
26,2
5
13,8
18,4
8,44
8,32
12,4
10
5,3
5,2
27,1
25,3
4,4
15
3,0
3,3
68,9
66,6
1,95
20
3,5
2,3
118
143
1,40
25
0,4
1,1
180
-
0,70
32
0,35
Mean values of the potential gradient dVids and of the con-
ductivityl4. in the free atmosphere published by Lug�on
et al. (1) and Muhleisen und Fischer (2)
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
To the global events there isacontribution by Fischer H.J.
The atmospherii; electrical potential gradient between earth surface
and about 15 km above has been measured continuously by more than
50 balloon ascents. The ionosphere potential has been found out by
integration of the altitude. The mean value was 282 kV in agreement
with other authors. The precision of the measurements has been + 8 %.
The diurnal and annual courses of potential mean values agree very
well with the mean values of the Carnegie measurements. But the
single values fluctuate up to + 30 % of the mean-values. The same
fluctuation can be found in the Carnegie mean values of single days.
Therefore it seems not to be correct to compare single diurnal
courses of stations on the continent with the mean diurnal courses
on the oceans like it is sometimes done.
14)
Fischer has found a seasonal maximum in the course of the ionosphere
potential during the northern winter. He explains this in agreement
with Whipple (1929) by a much larger electrical efficiency of the
tropical thunderstorms. Considering all thunderstorms on the earth
with the same importance, they would have their maximum of activity
during the northern summer and no accordance with the mean results
would be. If one gives however the tropical thunderstorms a greater
weight than the other thunderstorms on the earth, a special explanation
of the newest isobronten demonetization of the W110 material distributed
to different degrees of latitude proves that the maximum of the
thunderstorm activity is during the northern winter, indeed.
3) Conductivity, ions. radioactivity.
A great miner of publications has been submitted on the subject
of conductivity.
Bricard J. 15) improves his former theories about the
combination of small ions at aerosol particles. In calculation of
the coefficient in the formula for the ionization equilibrium he
takes in consideration, that the diffusion is not taking the normal
approved for Release: 2017/09/11 C06461858
,
Approved for Release: 2017/09/11 C06461858
course, if the particle radii have the same dimension as the mean
free path. For the radii more than 10-5 cm the deviation can be
neglected. It gets new values for the combination coefficient in
the radii range 0,6 z 10-6 to 4 z 10-6 cm, which have a stronger
deviation as the former ones, but they agree well with the results
of Keefe, Nolan, and Rich.
Ninet's work (see 5 in chapter 1) concernsalso the conditions in
the ionconcentration near the ground.
Besides others Reiter R. 16)has registrated the conductivity
and the concentration of small ions at some places of his atmospheric
electricity stations at the Zugspitze. Remarkable but not quite
clear in the fact that the ratio n+/n_ shall increase with the
speed of the wind: For vw 0 he finds n+/n_ = 1,0; for vw 5m/s
he finds n+in....7 2,0.
Occasionaly he finds a strong increase of the negative conductivity
of the air at the sun radiated mist and fog layers. He supposes that
it is caused by a photoemmission by the solar UV. It does not seem
to be correct, because there is no light of sufficient short wave
length in the altitude of 3000 m.
There also appeared some new publications in the subject of measuring
17)
technique. Hock A. and Schneer H. describe a new
small-ion counter, where they have disposed the counterfield effect
by grounding the aspiration-condensator coat and by putting back the
electrodes Antothe cylinder, where they have used double-eleetrometer
valves in the entrance of the direct-current amplifier.
An interesting new method about the direct registration of the
atmospheric small-ion spectrum has been published by Junod A.,
18)
Sanger R., and Thams J. The authors used a
measuring condenser with a linear, quickly increasing voltage, where
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
the currents become compensated by a bridge circuit. Not only the
current voltage characterism will be won, but also a mobility-
spectrum by a thoughtful use of electronics. It is a great advantage,
that all signals exist as alternating voltages and therefore the
direct-current amplifier can be avoided.
19)
Basencle�er D. and Siegmann H. published
a new method of dust measurement using small ion dissipation. In a
ionization chamber, working in the range of satisfaction, small
ions will be produced with a radioactive probe. The measured
ionization current changes when dusty air enters the chamber. The
small ions become partly combined with the dust particles. These
charged dust particles will not be measured for their lessened
mobility; the decreased current is a measure for the concentration
of dust. The authors show that not the dust concentration, but
the product of dust concentration and medium particle radius has
been measured. because of the dependence of the dissipation
coefficient on the radius of the particles.
Sik�na R. 20)developed an aerosol measuring instrument at
the same time, which works with the same system. Here the small
ions will be produced in a separate tritium-ion generator and
mixed with air in a special chamber. The content of small ions
in the air will be measured with an aspiration condenser. The
pecularity is that the production of small ions, the mixing of
the air, and the measuring of small ions is separated. The time
of mixing is also independent of the speed of the wind and given
by the speed of the air flow through the measuring condenser.
Both arrangements can be calibrated only limited. The increase
of small ions depends on the dissipation coefficient and this
. depends on the aerosol spectrum. The aerosol spectrum has a
further influence on the result because the time up to the
equilibrium is dependent also on the dissipation coefficient.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
The mixing time at disposial has to be more than the time constant.
Comparing measurements with the same aerosol spectrum can be made
with this lethod, where it will also be the advantage of a continuous
measuring. qn spite of all it is not satisfying, because there will
be kept only a single information.
Siksna R. and Lindsay R. 21)developed a small-ion
generator with a tritium soufce in a titan foil for the above
mentioned instrument. Herewith they have been able to place in a
small room a large activity and to produce a small ion concentration
up to 5 x 106 per cm3 of air. Of great advantage is the half-value
time of the tritium of about 12 years and the assurance that no
unwanted changes will happen with the aerosol.
There are also some works of interest on the subject of atmospheric
radioactivity. They have only been mentioned if they are in connection
with atmospheric electrical problems, such as the air conductivity
near the ground, the electrode effect or the atmospheric exchange.
Budde E. and Israel H. 22)discussed the diffusion
coefficient of the radon in the air in soil. H. Israel compared the
exhalation and the radonOoncentration in the lower atmosphere calculated
with the various values of the diffusion coefficient with the measured
values of these quantities. Be receives the result that a value of
,
D In about 0,05 cm2/sec approaches best to the actual condition. This
value depending on the sort of soil is about 50 to 500 times as
high as the one of Budde.
Lugeon J., Junod A., Wasserfallen P.,
Rieker J. 23) registrated the radioactive content of the air
near the ground besides the different qualities of atmospheric
electricity in Payerne as well as at the Murchison-Bay (Spitsbergen).
4
Reiter R. 24)displayed his results of some of his investigations
4
s of the natural and the artificial radioactivity measured in his two
.1;
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
stations Wank and Farchant. Because these two stations have a
difference in level of 1,1 km at a relative little difference of
base, the author can make some declarations about the dependence
of the components on the altitude. Out of it he derives the
influences of the meteorological parametemon the natural and
artificial radioactivity. He determines the exchange coefficient
by the calculated half-value altitude and he compares it with the
temperature gradient. These relations cannot be taken as the
general, because they are based on the conditions in a tight valley
of the mountains.
Ernst F., Preining O., and Sek_dlacek M.25)
investigated the size distribution of the radioactive particles in
the atmosphere of Vienna by using a Goetz-aerosol-spectrometer.
The filter has been tested by the autoradiographic method, later
the filter cut in single pieces has been investigated by a counter.
They make the conclusion that thegresbUtactivity of particles can be
found in the area of less than 0,7/u Stokes' radius.
Bricard H., Pradel J., Renauz A.26) employ
their dissipation coefficient to work out the frequency of the
dissipation of radioactive small ions on the aerosol particles of
different radii. For that the size distribution by Junge has been
used. It has been supposed that the RaA atoms caused by decay of
radon, fora small ions which combine later with the aerosol. Using
the formula the distribution of the activity on the aerosol particles
of different sizes can be worked out. The results have been com-
pared with the values also measured by the authors.
%) Precipitation electricity and electrification.
27) .
Reiter R. investigated the frequency of the eigne of the
potential gradient in his stations in the mountains of Wetterstein.
By a statistic exploitation of the spacious measuring material he
MIMApproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
found out the following proportions of signs during precipitation:
Table 2
Type of precipitation: rain rain shower snow shower snow fall
pos. signs of the PG:
10%
30-40
%
40-50
%
80 %
neg. N s s
90%
60-70
%
50-60
%
20 %
change per hour:
0,8
2,5-3
2-2,5
0,9
He analysed the number of the changes of signs per hour in the various
forms of precipitation in a similar way and he finds out the results
in table 2.
It seems to be very interesting that the changes of signs are much
more numerous in the valley than in higher altitudes. In the same
work the dependence of the sign of the potential gradient on the
altitude during steady precipitation in the two phases has been .
discussed. Above the melting zone the potential gradient is about
3 - 4 times greater than the value of fine weather, while it is
negative and about 2/3 of the value of fine weather beneath the
melting zone.
5) Numerous and valuable work has been made about thunderstorm and
thunderstorm theories, lightning phenomena and their electro-
magnetic signals stories and whistlers.
W o 1 f P. 28) discusses the present ideas of the cause ef
thunderstorms and the lightning-formation. Teeplee, ideas about
the formation of discharges have been put in the forefront again.
Pu 29)hringer A. put up for discussion a new thunder-
storm thierY , based on the electro-magnetic induction. In the
author's opinion an electrical field will be induced by the
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
motien of a cloud in the magnetic field of the earth. This produces
ea electrical dipol moment which shall build up a strong electrical
field in the outer room. The efficiency, necessary for a thunder-
storm, shall be withdrawn from the cinetic energy of the wind.
Michh�W 30) sk1 St. indicates the important part of the
point discharge in the preservation of the charge exchange between
the earth, thunder-cleuds and ionosphere. Measurements of the sum
of the point discharge for � longer period in Swidrze gave a
result of the mean ratio of a charge run out of a point gs 1,5.
Mi 31)l ler-Hil lebrand D. describes some iatereeting
thunderetorm observations at Monte San Salvatore. The value of the
electric field has formerly been estimated from about 20 kV/m
to 330 kV/n. The author noticed only a field-strength of about
3-5 kV/m near lightning-strokes in the ground. He tries to ex-
plain whereby the strong fields have been screened by a larger
space-charge area; registrations of the electric field, point
discharge, precipitation current and precipitation strength ot
the same time support this assunptien. An exact temporal analysis
of the formation and discharge of a flash had been possible by
measuring of the lightning current at the place of stroke and
the electrical and magnetic field-strength in a distance of
2,8 km at the MIMS time. It had been shown that the steps
of a leader stroke are extrenty short. The single impulses
can have a temporal interval down to 0,2/u sec.
the
In another publication the author tries to extend the protective
radius of lightning conductors by radioactive point discharges.
At the approaching of a thundercloud, all points emit some
corona currents, which will be led away by the wind as space
charge. It had been asserted radioactive poinV could increase
the point discharge current so much that the space charge cloud would
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
a
lead to catch discharge. La a laboratory experiment there was no
difference noticed concerning the corona current of a radioactive
point and a normal point after arising the corona discharge.
With another experiment it was however shown that a strong point
discharge of 8 mA has a strong influence on the electrical field
strength in the surrounding.
Outdoors the point discharge current of the radioactive points
was always less than a current of non radioactive points. Obviensly
the ion cloud around the radioactive point delays the beginning of
the corona discharge.
32)
HUI ler-Hil lebrand D. has won an interesting
material about lightning frequency since 1958. With a rather
narrow net of stations (115 stations) in Sweden he was able to
registrate more than 100.000 earth flashes in 1958. These made
65 % of all registered discharges.
Malkowski G. 33) determines the mean value of the diameter
of a convective precipitation cell d 4 km by a collective of
1000radar observations during showers and thunderstorms. He gives
a curve for the frequency distribution of these values. At the
observations of echos with the weather radar instrument it is of
some interest whether a precipitation cell visible on the radar
screen can be considered as a thunderstorm or a shower without
lightnings. This had been undertaken by a sferic direction finding
instrument by finding the position of thunderstorm centers at
the same time.
34)
Norinder H., Knudson E. made spacious
investigations about the discharge mechanism of lightnings in
a free field station near Uppsala (Sweden). The collected data
had been exploited. The length of the lightning path between
cloud base and earth gave values of 0,6 - 2,4 km with a mean
F.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
value of 1,4 km. Observating 1135 flashes, there were 79 % between
clouds and earth and 21 % within the clouds. Analysing the multiple
discharges it had been displayed a decrease of the magnetic field-
strength with the number of strokes. The intervals between the
lightnings had been noticed from 16 thunderstorms. It was found,
that the most intervals are 20 - 70 sec. At a single thunderstorm
of extreme strength 80 % of the intervals had been shorter than
1 minute.
35)
Berger K, made lightning obersations with an oszillograph
on the Monte San Salvatore in 1958-61. The registrations had been
exploited concerning the front duration, the maximum current, the
average and maximum current increase slope. On the base of this
measurements there had been 4 different types of lightning discharges,
which differ in front duration, maximum current, steepness of
increase and current curve:
a) flashes with leader strokes in upward direction; in general they
only come from high and well grounded conductors. The form of current
shows a slowly increasing with a current maximum of 20 to some
100 A;
b) flashes with stepped leader from a negative cloud; the main
discharge has a maximum current of 15 - 45 kA and a front duration
of 4 - 12/u sec;
c) following strokes from a negative cloud. The maximum current
strength is smaller. The front duration is shorter and less than
lin sec. in general;
d) discharges from positive clouds to the earth with a slow current
increase and a high discharge strength. A discharge exchange of
more than 100 C is nf no rarity.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
r
P7r-z
36)
Norinder H., Knudson
made experimental investigations about the multiple-lightning
strokes in the same lightning channel in 1956-57. Multiple
lightning strokes have been regarded in the laboratory if a high
ohm-resistance has been inserted in the discharge circuit. Because
multiple lightning strokes can give some indications on the
discharge mechanism in the thundercloud, oszillographic registrations
of the strokes series and the discharge currents of natural
multiple strokes. In order to registrate the front time and the
discharge current three oszillographs with different time bases
have been used. The electromagnetic field had been received from
a frame areal and led to the oszillograph by an aperiedical
amplifier. Herewith the changes of the magnetic-field strength
had been registrated and an estimate of the current changes in
the lightning channel has been possible. The measuring results
had statisticly been exploited concerning the amplitude dis-
persion, front time and temporal intervals. Later on the method
had been completed by day-light photography. The increase of
voltage in the aeral triggered a connection circuit, which
released the camera. This combined method made it possible
to declare something about the discharge process in the cloud.
Papbt Lpine J. 37) has been occupied with a
theoretical work about the mechanism of the lightning discharge
and the herewith existing change of the magnetic field. The
author shows a method to calculate the temporal course of
current in the lightning channel from the magnetic field
changes measured near lightning discharges.
38)
Fritsch V. has been occupied with the problem of
geological and geoelectrical influences en the place of
lightning-strikes. He can confirm the opinion, that there
are lightning nests. A special result was, that the danger
by lightnings grows with increasing geolotiul age of the
terrain.
...F. � 1.'
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
39)
Michnowski St. describes an interesting observation
in north:trn Viet-Nam, where electrical discharges had been in a
cumulus cloud without any signs of existing ice crystalls in the
cloud. The altitude of the upper cloud limit was supposed to be
2500 m, while the 00 isothermefor this season should have been
about 4000 a.
De�sens H. and his coworkers40) made investigations with
an installation of 100 burners which are arranged on a quadrat of
a sidelength of 100 m. They tried to produce artificial cumulus
clouds. During the first few experiments there havebeen developed
sometimes tornado pipes with wind speeds of some hundreds of km/h.
It had been noticed that lightning discharges ensue always along
natural tornado pipes. Deesens hopes that aninstalled ventilator
will stimulate the formation of a whirl of a tornado pipe. Hessen'
expects that the so called Meteotron gives him the possibility
for direct lightning investigations and for experiments in plasma-
physics.
An earlier theoretical publication by M.O. Schumann refers to a
resonnunce.frequency of the condenser ionosphere. earth. This
41)
frequency shall be about 9 c/sec. Honig H. investigated
there upon atmospherics of this extremdy low frequency. His receiver
connected with a long-wire antenna enclosed the range from 0,5 - 13.c /sec.
In fact he got signals with frequencies 8 - 9 c/sec. He believes that
one part of these signals is caused by lightnings, which have excited
the resonnance circuit ionosphere earth to oscillations. Another
part of signals arises at sun rise. Kiinig supposes, these signals
would be caused by abruptalganges of the altitudes of the lower
boundary of the ionosphere. Other types of signals with lower
frequencies will be brought into correlation with local weather-
phenomena. Honig has examined his interesting results by a
simultaneous registration: .on a second station in a distance of
50 kmresp.450 km on one hand,on the other hand during the eclipse of
the sun on 11 15.2.1961 (E.Haine).
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Malkowski G. '2) determines the entrance-range of a sferics
receiver by a decrease of the sferics frequency during a antizyclonic
situation. At a field strength limit of 0,4 Vim he comes to a range
of 850 km.
43)
Israel H. compares the sunrise effect of sferics at 27 kc/sec
registrated at 3 different stations, Aachen 50,8� N, Tokyo 36� N,
San Salvador 13,5� N. The beginning of the decrease of the sferics
intensity varies from station to station during the annual coarse.
The nearer the station is to the equator the earlier and more unregularly
the effect begins.
44)
Vattern G. delt with the reception of sferics from great
distances. He had built up his station on a solitary island without foreign
electric installations. He is able to receive all lightning signals
from the whole earth. This
however leads to overlapping signals
so that a lower limit of sensibility (about 50/uV)must be fixed.
Norinder H. and Knudenn E. 45) investigated
whistlers 1957 - 59 in order to explain the connection of thunderstorm
activity and whistlers. It has been proved that only one part of
flashes produces whistlers. They septly appear in groups for a time
of 1/2 to 2 1/2 hours. The longest time of observation had been
6 hours. These times favourable for whistlers have been intelUpted
by long period* of silence. Registrations of sferics at the same
time showed that only stories with the highest field-strength have
been followed by whistlers. The exploitation of the field course of
the stories with a harmonious analysator gave as a result that the
energy maximum of the radiation is on about 5 kc/sec. A comparison
between the temporal series of the multiple whistlers and the
oszillographic registrated multiple-lightning strokes shoved
accordance concerning the time intervals.
Different whistler shapes had been regarded by the analysis with a
"Sonagraph". One third of all registrated whistlers during the
pproved for Release: 2017/09/11 C06461858
SM�i�I
Approved for Release: 2017/09/11 C06461858
thunderstorm season 1959 could not have been put into relation of
thunderstorm- renters by a sferie-directionfinder. All these
whistlers appeared only during a short period and must have been
produced in the neighbourhood of the conjugate geomagnetic point.
46)
In another publication Norinder H. supposes that the
propagation ef whistlers will be influenced by ionospheric irregularities.
He follows the ideas of Budden, E.G. in a theoretical publication.
6. Space phenomena of atmospheric electricity
In this chapter only seme investigations can be reported which are
due to cosmic rays. The results come from balloon ascents made in
one institute.
. .1
47) 6
Waibe 1 E. determined the ionization spectrum of the cosmic L.
;,..
0-
-,, -
rays. He was able to separate the protons from He- and Li-nuclei
in the primary radiation. If one extrapolates to the limit of the
atmosphere, theoc,-intensity is about one seventh ef the proton
intensity.
r b � H. 48)finds a clear relation between the intensity in
definite altitude measured by a Geiger-counter telescope and the
intensity on the ground.
Ehmert A., Erb� H., Pf�ther G., Keppler E.
discussed the pecularity of commit/ rays in a solar eruption in
outman 1960. During the summer season 23 balloon ascents had been
made in Kiruna (Sweden). Some X-rays eruptions and injections of
solar protons have been observed.
Ehmert A. 50) indicates that the experimental "rigidity
spectres" of the primer cosmic protons arida.- particles can be
described by a variation of the electric potential of the earth
49)
� Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
against far space. Its variation is correlated with the solar
activity. The decelerating potential has a variation of 1 Gigavolt
at sunspot minimum and 2,7 Gigavolt during a magnetic storm. The
field is supposed to be beyond the radiation belt, as oaf can
conclude from the intensity observations of noon rockets.
7. Biological action of atmospheric electricity.
Knol 1 M., Itheinstein J., Leonard G.F.,
5l)
and Highberg P.W. investigated the influence of
artificial atmospheric small ions on the reaction time and the
visual moment. An increase as well as a decrease of 7 % of the
reaction time has been found for desities of about 103 to
106 ione/cm3, if the subject is breathing through the mouth.
There is no influence when the subject is breathing through the
nose. The polarity of ions does no matter at all. An influence
for the optical moment - i.�.. the shortest tine between two
flashes, which can be recorded separately - has not been found.
The influence of ions resembles the effect of many drugs en
the human system.
52)
K onig H. has registrated the atmospheric impulse
radiation since four years. The receivers are able to record all
sferics in three entrance ranges: 100, 300 and 1000 kn. Until
now there was not found any clear relation with aspects of
illness on human beings.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
R�f�r�no�e
f) Saxer, L., Signet, W.: Tages- sad Jahresgings der luftelektri-
schen Element, in Areas.
Arch.Met.Geeph.Biekl. 12 (1961) 366-75 Ser.A
2) Israel, H.: Meteorologische Vergange is Spiegel der Luft-
elektrizitat.
Annali di Geefleica 12 (1960) 327-367
3) Hilihicieen, S.: Latelektrische Verhaltaisse is Usteaaeresel I.
Arch.Met.Geoph.Biekl. 11 (1959) 93-108 Sen. A
Mthleleen, R.: Luftelektrische Verheltnisse LE Kfietenaeromol II.
Axch.Met.Geoph.Bioel. 12. (1962)
Milhleisen, R.: Electrode effect measurements above the sea.
J.Atm.Terr.Phys. 20 (1961) 79-80
5) Ninet, C.: lalation entre lee permtree 616ctriques do
l'atmesphire it la diffusion turbulente.
Dissertation Universit6 Paris 1961
6) Reiter, R.: Luftelektrisches Erecheinungebild des Sidfdhns in
den Nordalpen notch eynoptisch-klimatologiechen Ulster-
,m
suchungen in Wettersteingebirge.von 1955-1959.
Arch.Met.Geoph.Biokl. 12 (1960) 72-124 Ser.A
7) Kilineki, E.: Verbesserung der Auffangvorrichtung des Petadamer
Vertikalstromgeritee.
Z.Met. 12 (1958) 352-54
8) Israel, H.: in W.Hease Handbuch der Aerologic,
Leipzig (1961) 646-701
9) Georgil, W.: in Clime Handbuch der Aermlogie.
Leipzig (1961) 274-277
10) Mibleisen, R., Fischer, N.J.: Luftelektrische Aerologie.
Beitr.Phys.d.Ata. 2.�.� (1961) 3-14
11) Lagoon, J.: Electresendagee i deux conductibilitis pour is
ditectien du niveau de is vase atmesph6rique.
Act.Sec.ilelv.Nienne (1961) 93-95
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
0.
12) Lugeon, J., Junod, A., Wasserfallen, P., Rieker, J.:
Hesures des Parasites atmespairiques d'Electriciti
atmosphfirique et de Radioactivitio de l'air
Murchisen-Bay (Spitsberg), Payerne et Zurich.
Zurich 1960
13) Maleisen, R., Fischer, H.J.: liessung des luftelektrischen
Feldes in der freien Atmosphire.
Naturwiss. la (1960) 36-37
14) Fischer, H.J.: Die elektrische Spanning zwischen Ienosphire
und Erde.
Diss.Stuttgart 1962
15) Bricard, J.: La fixation des petits ions atmosphiriques cur
les atr000ls ultra-fins.
Geof.pura � Appl. (1962/1) 237-242
16) Reiter, R.: Relationships between atmospheric electric
phenomena and simultaneous meteorological conditions.
Final Report, Contract AF July 1960.
17) Hock, A., Schneer, H.: Uber em n Gerit sur stOrfeldfreien
Luftionennessung....
Z.f.angew.Phys. 14 (1962) 398-404
18) Junod, A., Sanger, R.; Thane, J.Ch.: Enregistrament direct du
spectre des petits ions atmesphiriques.
J.de Math.Phys.Appl. 12 (1962) 272-278
19) Hasenclever, D., Siegmann, H.: Nene Method, der Staubmessung
nittels Hleinionenanlagerung.
Staub 20 (1960) 212-218
20) Siksna,
21) Siksna,
22) Israel,
23) Lupton,
R.: An ionemetric counter for condensation nuclei.
Geilf.pura � Appl. a (1961) 23-36
B., Lindsay, R.: Air ions produced by a tritium-ion
generator.
Arkiv for Geofysik (1959) 123-154
H.: Der Diffusiens-Hoeffizient des Radons
Bodonluft.
Z.f.Geoph. (1961) 13-17
in
J., Juned, A., Wasserfallen, P.:
Meeures d'electricit6 atneephirique et de is
radieactivitit de l'air Murchison-Bay (Spitsbergen)
pendant l'AGI 1957-58
Acton de la Secilete Helv6tique des Sciences
Naturelles Lausanne (1959) 122-126
approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
24) Reiter, R.: Neu* Ergebaisse alpiner Luftradieaktivitilts-
messuagen. 4 1-
Zbl.f.biol.Aeioselfersch. 2 (1960) 3-27 h
,.
r...r,,.-
i1.7.
.w.; 25) Ernst, F., Primping, O., Sedlacek, M.:
Sizing of the Fall-out at Vienna.
Nature 122 (1962) 986-7 L 4.
26) Bricard, J., Pradel, J., Renauz, A.:
Le Linear de Pair en petits et gres ions radioactifs.
Goef.pura e ppl. 52 (1961) 235-242
'
27) Reiter, U.: see16
28) Ifelf, F.: Gewitter
Physalitt. iz (1961) 501-512
29) Puhringer, A.: Die elektromagnetische Induktion ale Grundlage
einer Gewittertheerie.
Arch.f.Mot.Geoph.Biekl. 12 (1961) 262-70 Ser.A
� 30) Miehaewski, St.: Point discharges in the interehange of
electric charge between the earth and the atmosphere.
iota Geoph.Polon. 2 (1957) 123-134
31) Mfiler-Hillebrand, D.: Beeinflussung der Blitzentladung.
ETZ A 82 (1961) 232-249
32) Miller-Hillebrand, D.: Lightning counter and results obtained
in Sweden during the thunderstorm period 1958.
TV?, toka.-vet.Forskn.22 (1959) 1-16
33) Malkowski, G.: her die Grosse konvektiver Niederschlagszellen
auf den Radarschira.
Geol.Beitr.z.Geeph. (1961) 51-57
34)
Winder, M., Knudsen, E.: Some features of thunderstorm
activity.
Arkiv for Goofys. 2 (1961) 367-374
35) Berger, K.: Freatzeit und Anstiegesteilheit des Blitzstremes
bei Erdblitzen.
FIB 11 Zurich October 1962
36) Winder, if., Knudsen, E.: Lightning discharge paths photo-
graphed in day-light and analysed by simultaneous
variations of magnetic field components.
Ark.for Goof. (1961) 375-390
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
37) Papet-Ltpine, J.: Electromagnetic radiation and physical structure
of lightning discharges.
Arkiv for Geoph.2 (1961) 391-400
Papet-Ltpine, J.: Rayonement electremplomtique et structure
interne des eclairs.
Dissertation Univereitt) de Paris bet. 1959
38) Fritsch, V.: Geophysikalische Einflosse auf die Blitzgefahrdung.
Scientia 2i (1960) 1-7
39) Michnowski, St.: On the observation of lightning in warm clouds.
Private communication.
40) Dossing, H.: Un physicien francais cherche a canaliser
in toudre.
Le Monde, 27.Juli 1962
41) Honig, H.: Atmosferics geringster Fregnenc.
Z.f.angew.Phys. 11 (1959) 264-274
Honig, H., Hain., E.: Registrierung besonders niederfrequenter
elektrischer Signale wahrend der Sonnenfinsternis am
15.2.1961.
Z.f.angew.Phys.12 (1961) 264-74 and 478-480
42) Malkewski, G.: Zulu Einsugsbereich eines Sferic:-Empfanger.
Met.Rundsch. 12 (1959) 64
43) Israel, H.: On the sun rise effect of sferies activity
at 27 ke -I.
Z.f.GeoPh.26 (1960) 138-143
44)
Mattern, G.: private communications.
45) Nerinder, H., Knudsen, E.: Lightning discharges as a genres
of whistlers.
Ark.for Goof. 2 (1960) 255-288
Nerinder, Up;. Knudsen, E.: Multiple lightning discharges
followed by whistlers.
Ark.for Gee!. 2 (1960) 289-298
Nerinder, H., Knudsen, E.: Occurrence of different kinds of
whistler activity.
Arkaiir Goof. 2 (1961) 347-366
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
411.�
�
46) Norinder, H.: Sone consents on the penetration of whistlers
through ionospheric layers.
Planet Space Sci 2 (1960) 261-262
47) Weibel, E.: Messungen von Prinirteilchen der kosnischen
Strahlung.
Mitt.a.d.MPI f.Aerononie 1 (1959)
48) Erbe, H.: Auswirkungeo der Variationen der primaren kosnischea
Strahlnng ant die Mesonen- und'binklsonenkosponente
an Erdhoden.
Mitt.a.d.MPI f.Aeronomie 2 (1959)
49) Pfotzer, G., Ehmert, A., Erbe, H., Keppler, E.:
X-Ray Hunts in the Auroral zone on Sept. 27th0ctober let
and 2nd 1960. Space Research II.
Proc.of the 2nd Int.Space Science Sympalorence
April 1961, 876-886
Ehmert, A., Erbe, H., Pfotzer, G.:
Pecnlarilies of the outburst of solar high
energie particles on Nov.1960. Space Research II.
Proc.of the 2nd Int.Space Science Symp. Florence
April 1961, 778-786
50) Ehmert, A.: Electric field modulation.
Space Research I.
Proc.of the lit Int.Space Science Symp. Nice,
Jan. 1960, 1000-1008
51) Knoll, M., Rheinstein, J., Leonard, G.F., Highberg, P.W.:
Influence of Light Atnospheric Ions on Human Visual
Reactien Ti...
Proc.ef Int.Conf.on Isnization of the Air, Oct.1961
18, 1-25
Rheinstsin, J.: Der Einfluss von kiinstl.erzengten atm. linen
ant die einfache Reaktionszeit and ant den eptischen
Resent.
Dissertation TB Munchen (1960)
52) Kbnig, H.: Atnospharisch-elektrische Untersuchungen on
Meteorologischen Observaterinn Hamburg.
DWD Hamburg
pproved for Release. 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
� Atmospheric Electricity Research*
Great Britain, Ireland, Africa and Nctw Zeeland.
4: Point Discharge and Precipitation Curmnts
pproved for Release: 2017/09/11 C06461858
pproved for Release: 2017/09/11 C06461858
The purpose of this paper is to present a survey of research in
Atmospheric Electricity performed in the countries concerned and
reported since the Second Conference on Atmospheric Electricity in 1958.
The survey is not intended as an index to all the relevant literature,
nor is the space allotted to any tern to be regarded as a measure of its
Importance. I have arbitrarily left out work on atmospherics and
electrcmagnetic wave propagation. To make the picture more complete it
is well to refer to the present quite conaideroble and wide-spread interest
in the subject of Atmospheric Electricity In these countries.
I will mention five centres in England. At Lcutherhesd the Electrical
Research Association with over 1000 voluntary observers has since 1950
collected data on thunderstorms in Britain. At. Imperial College, London,
the experimental and theoretical studies of the electrical properties of
ice and water are proceeding in Professor Meson's Sub-Department, and the
work of Dr. Browning and Dr. Ludlam (1962) on the airflow in convective
storms may well lead to a new approach to the problem of thundercloud
electrification. Dr. Wormell's group at Cambridge University are extending
their investigations on ions and Aitken nuclei, and arc also studying low
frequency fluctuations in the earth's electric field and the field spectrum
of near lightning flashes. Dr. Latham, now at the Manchester College of
Science and Technology, is continuing his studies on the frictional elec-
trification of ice. At Durham University Dr. Chalmers' group are investi-
gating precipitation and air-earth currents with mobile an well as static
apparatus, point discharge, space charge, and electrical effects associated
with wetter and ice. In New Zealand, at Auckland University, Professor
Kreielshetmer is concerned with potential gradient and point discharge effects
at balloon altitudes. There are reports from three centres in Ireland.
Professor Pollak and his group at the Dublin Institute for Advanced Studies �
are working on the electrical equilibrium of aerosols and ice-nucleus con-
centration determinations, and have constructed a small portable photoelectric
nucleus counter. At University College, Dublin, under Professor Nolan,
research is on charged and uncharged nuclei and the application of the
Boltzmann Law to earlier observations. Dr. O'Connor and his group at
pproved for Release: 2017/09/11 C06461858
� Approved for Release: 2017/09/11 C06461858
j
V
2.
cellep.., *away, tire "ar.,w4 In Si nIqdy "r isoAtmt
Altittn vuelei and space charge.
The African %:ontinent, with its fine opportunities for research in
tropical and sub-tropical regions, now has several centres. In South
Africa, at the Bernard Price Institute of Geophysical Research,
Johannesburg, work under Dr. Malan includes the study of lightning
flashes uf various :ynes, lightning photography. the! study of upward
dischnrges above clouds, and flash comaing techniques. Also in South
Africa, in the Univtruity of Natal, at Durban, Professor Clerence's
Department is containuing research on whistlers and lightning. Observations
are now being made at the University College of Sierra Leone, at Fourah Bay,
of point discharge and precipitation currents. At University College,
Ibadan, point discharge currents arc being studied, and an interesting
investigation of lightning by sound-ranging on the thunder has also begun.
Other centres in Africa where there have been projects for Atmospheric
Electricity research are at Salisbury, Southern Rhodesia, Mftkarere University
College, Uganda, and University of Nigeria, Naukka.
2. Ionization in the Atmosphere
The time required for a cloud of uncharged nuclei to reach equilibrium
has been further investigated by Rich(1), Pollak(2) and Metnieks(2) (1962).
Their calculations involve integrating the equations for the rate of change
of concentration of small ions and charged nuclei respectively:
dn
q - - no n No - A n mr, and
qo n No - A n H.
dt a
Here n, N are the concentrations respectively of small ions and charged
nuclei, assuming equal numbers for either sign, and No the concentration of
uncharged nuclei. The number of small ion pairs produced per ce per sec is
q, and A, qo are the appropriate combination coefficients. /t is assumed that
(1) General Electric Couiptuky, Schenectady, U.S.A.
(2) Dublin Institute for Advanced Studies, Ireland.
� :��
'It, .11,
�
r � .!rs
�.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
3.
multiple charging or recasbination of nuclei muy be neglected. The
authors started with initial values N a 104 cm-8 and n a 950 cm -a and
��� � !.
o
o; �
qv. 1.6 ad-8 sec l and considered three different values for the fraction
�
of nuclei charged at equilibrium, corresponding to three different values
of nucleus radius. The results of the calculation show how much more
slowly the charged nuclei approach their equilibrium concentration compared
with the mnall tons, except for very small nuclei. With their arbitrary
initial conditions the authoru found that for nuclei of radius 3.6 x 10-8 cm
n is at its equilibrium value after 7 minutes whereas N takes nearly an hour
to reach 90% of its equilibrium value. These results together with the
recent history of an air mass may be used to estimate whether the nuclei
in it are in charge equilibrium. In one observation, however, upwind of a
town, and where charge equilibrium might have been expected, the concentra-
tion of charged nuclei was only 0.6 to 0.8 of its equilibrium value, perhaps
of undetected sources of nuclei.
Pollak und Metnieks(3) (1962) measured the rate at which a stored
aerosol approaches charge equilibrium. MAORI of various sizes were pro-
duced by heating a niehrome resistance-wire inside a 4.2 m8 balloon. At
intervals during the decay of the resulting aerosol they took a sample
and measured the traction of the nuclei charged. Simultaneously another
sample was brought to charge equilibrium, using a wenk Oc-ray source, and
the ftmetiou of charged nuclei measured. A ante or charge equilibrium
was recognized when the fractions for the two :maples were equal. For
stored euclei or equivalent radius 3 x 10-8 cm androncentration 22,000 cd-8
equilibrium was reached witL.n 15 mln. Au the size and concentration increased
so did the time taken. Nuclei of radius 10-8 cm and in concentration falling
with time from 234 x 108 to 59 x 108 ad-S required several hours. The largest
had not reached it even after several days. These results confirm the
experimental predictions described above.
(3) Dublin Institute for Advanced Studies, Ireland.
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
The use of an a-source in bringing an aerosol to charge equilibrium
has been investigated theoretically and experimentally with a stored
aerosol by Flanagan and O'Cetnor(4) (1961). They conclude that it provides
the best method at present avalluble to test for charge equilibrium.
Following work briefly reported by Nolan(5) (19')5) at the First
Conference on Atmospheric Electricity the problem of the equilibrium
concentrations of charged and uncharged nuclei in air has been further
examined by Keefe(6), Nolan(6) and Rich(T) (1959) by applying the
Boltzmann Distribution Law, assuming that becauue of their frequent
collisions with small ions the particles should be in charge as well as
in thermal equilibrium. A charged particle carrying x electronic charges
C is treated as u spherical conductor of radius r so that it has electrical
ener x9e9/r in addition to the energy E0 of an uncharged particle. The
By Boltzmann's Law the number in unit volume N(E) having energy E is given
where A is a constant and g(E) is the statistical weight of the energy state.
E. Since a particle has the same energy whether its charge Is positive or
negative the utatintical weight of the energy ntate x >0 is fix m 2. Hence
the number per unit volume with x elementary charges regardless of sign is
where No is the concentration of uncharged particles and Ix' >0. If the
numbers of positive and negative partitles are equal, the number per unit
volume carrying x elementary charges of one sign is *lex. Writing y m e2/rkT
the total number of charged particles of one sign in unit volume is given by
(4) University College, Galwuy, Ireland
.(5) University College, Dublin, Ireland
(6) University College, Dublin, irelftud
(1) General Electric Campany, Sehenertady, U.S.A.
Approved for Release: 2017/09/11 C06461858
5.
NiNo e-Y + e-hY +e-9Y +....
- + li(15 + s-R2/Y + c-62/Y + e-9112/Y + ...) - y c
The latter form is more convenient for the larger particles, say r > 2 x 1(P5 cm,
when all the exponential terms arP negligible compared with so that
WN0 = + Ash)
When r> 2 x 10-5 an, and if the total concentration Z a No + 2N, then
Z/No a Ash)
.1(21trk1ic2) �
KA. where K Es constant.
The values of Z/No so dednced are In taxa ngreement with the observtaions
of Nolan and Kennun (1949) cmi the equilibrium charge distribution or
nuclei, derived from hot platinum, in the size range o.T x ].Cr5 < r <
14 x 10-5 cm.
It is further shown that for the larger radii the Boltzmann Law
treatment predicts an average charge per particle of ./(2rkT/m). For cloud
droplets of radius 5 x 10-4 an this would give a specific charge of 6.8
gO-1. The average electrical energy per 'article is shown to be ikT,
the value to be expected frau the classical law or equipartltion or energy
If the charge on a purttele in regitrikod an a eoordinute for one degree of
freedom, the en.�1.7 proportiontil to tilt, :;,ituire or the cluirgo.
Keefe, Nolan and Bich then upply the lioltzmattnIaw tom neronol In clutrw
equilibrium to deduee the ration the vnrIoun eembitation Poeffielents for
ions and nueleieWheu r >10-5 cm, but nut for the outlier particles, the
values of these ratios agree well with those deduced frcm earlier formulae
based on diffusion of ions and ionic mobility - the "diffusion-mobiliti form-
ulae".
An experimental investigation of ionization equilibrium In maritime air
has been made by O'Connor and Mara:y(8) (1910) on the west eonst of Ireland.
Upwind of the site there was no source of man-made nuclei within 15 km and no
major source within 1500 km. Assuming that the Boltzmann law applies,
(8) University College, Galway, Ireland
pproved for Release: 2017/09/11 C06461858
ift
Approved for Release: 2017/09/11 C06461858
6.
particles of a givee size should have a definite fraction of their numLer
charged. �.)",!ostaor and Sharkey determined the radii r of nuclei in the sea
airs:ram the diffusion coefficient found using a diffusion box and photo-
electric nucleus counters. They also measured the total nucleus concen-
tration Z and that of uncharged nuclei No. A graph of Z/No plotted
against r 4howed general but by no means eamplete agreement with the
theoretical curve Lased on the Boltzmann law. The authors note particu-
larly the Prevent large and erratic eluctuations in Z when its average
value wan high, and claim that on theue oceaeions equilibrium studies
were hapractiennle except by encle:Ing it large :ample th A 01:4JMOAq's
They dia.:ova the ye:Wile lack of equilibrium, Itte in the Introit:Ion of
� uncletrged fr.la eatural ti %I on the :get ehore.
Kecte anti tiolan(9) (:!t1 V.H32) have roc,geated a model for the capture
of small ions by uncharged euelei. When r 410-7 cm the combination is
assumed to be due sainly to eimple kinetic theory collision effects with
the effective target cross-sectional area increased by u factor due to
electrical Image forces. For large nuclei, when r 10-5 cm, diffusion
effects are predominant. In the lett:mediate range with r about, ltra cm,
t13 in the air at aett level, eapture is thuaett to he ante jointly to both
mechanisms. The authors caleulated the eombinatlon coefficient for Iona
and uncharged htleivi and found moet.rately good agreement with values
observed, but they emphasize the lack of good experimental data.
A study of nucleus and ion concentrations has also been made at
Cambridge. The work, by Adkins and by Law(1�, is referr.d to later.
5. Potdatt Oil Gradient and :kW., Chit elp�::
The fact that local eyncentrations or apttee eharge often seriously
modify the electric riad near the ground In all weather eonditlon.:, but
particularly in dilAurbed m4ither, has been siisphanized by AdkinaCu) (19./t).
He made continuous records of potential gradient with a field mill, of dptice
charge concentration using a steel wool filter connected to a vibrating reed
(9) University College, Dublin, Ireland
(103 Cambridge University, England
(11 Cambridge University, England
r:���
7fir
pproved for Release. 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858 .
7.
electrometer, and of small ion concentration with an Ebert ion counter.
In fine weather he found large fluctuations in Space charge Concentration
vhith made its mean value difficult to estimate, though it would probably
be about +2 pC m73. On several occasions in undisturbed weather he noted
a close correspondence between the records of space charge and potential
gradient. Sometimes this was associated with exhaust smoke from passing
traffic; sometimes the records followed a similar course for an hour or
more., usually in quiet, stable conditions, when the observed space charge
concentration changes would need to reach up stime tens of metres to acceunt
for the Observed field veriutions. From his mettsurements in disturbed weather
Adkins finds evidence for four processes. These ure
(a) the electrode effect (Aid) he discussen In detail),
(b) the modification or the potential gradient within some tens of metres
or the ground by large-ion space charges resulting from point discharge,
the small ion density remaining almost unaffected,
(c) in heavy rain, modification of the potential gradient by space charges
of smell ions produced by splashing (Adkins reproduced this effect in a
laboratory study and showed that the charge is proportional to the
existing field), and
(d) the control of the potential grndient near the ground by regions) or high
space charge associated with a column or min.
Adkins discusses the effective current due to splashing both in steady rain
and in heavy rain.
Law(12)
(1961a,b) has developed an automatic condensation nucleus
counter operating a pen recorder to study the vertical distribution of
nuclei within 3 m or the ground in connexion with stadiee of space charge
concentration. His unpublished work shown that convection plays an Impor-
tant part in the vertical transfer or electric charge.
The space charge concentration near the ground has been deduced by
13) -
Smiddy and Chalmers( 0.959, from measuremente at two heights using
Smiddy double field mills to minimize field distortion. In fair weather a
(12) Cambridge University, England
(13) Durham University, England
����
�
44.
..t ;�.:3
."
fa3
,
�:��t�
. � I
A .
ee '''.; �.1
' 4,..t.'es �
*4 l�f-; �
..;
,ra
;
� '4..7
.1;
.e.
..� sir 4ff�Ast-.��
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
8.
small negative space charge observed is explained in terms of radioactivity
of the ground, and, in heavy ruin, concentrations of negative charge up to
1000 pC m�s are reported. The authors suggest that the luck of agreement
with their simultaneous measurements using Obolensky filtration apTaratun
is due to the presence of small Ions.
Following the construction by Stein(14) (1958) of a field mill to be
carried by a radiosonde balloon, an ingenious double field mill for
rudiosohde work. big hen been de:Agned and made by Currin 4nd Krelelshelmer(15)
(196C). The stator and rotor memi�ers each comprise the oppwite quadrant.;
of a circular plate. With the two stators eunnect*.d togeth,:r the output In
proportional to the mill self-charge if the two rotor:: move in phnue but pro-
portional to the external potential gradient if the rotors maintain u relative
displacement of 900. Errors in field meusurement due to the charge on the
Instrument are thus automatically eliminated. The device has now been
prepared for carriage in a glider of the Imperial College Gliding Club for
thunderstorm investigations in England.
Adamson(16) (1960) has designed a field mill with overall negative
feedback giving a very closely linear relation between output and fiele
to be uned in :oniunction with tin unshielded air-enrth or rain eurrent
continuously recording system. 73o* mill output tg r.gt via a dift't-rentlutIng
circuit into the current amplifier in such a way UG to give compensation for
� �
A
e.
the displacement current which is proportional to the rate of change of field. j'"
V � �
The apparatus has a time constant of 20 secand excepting thunderstorms it is
� �
suitable for all weathers. i s � �
t
Wildman(17) (1962) has devised a field mill suitable for use when the �e:"/
0.
signal due to the conduction current is no longer Laval compared with the :
induction sigma. Nis machine rotor hun two eoneentric rigo or heleu ::�
���:?*
covering and uncovering two sets or innuluted etude, giving two neparatc
signals with different dependence on field and conduction current, allowing
the effects of these two to be distinguished.
(14) Auckland University, New Zealand
(15) Auckland University, New Zealand
(16) Durham University, England
(17) Durham University, England
proved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
9.
4. Point Discharge and Precipitation Currents
A new method of measuring point discharge current &Ilona tree has
been introduced by Maund and Chalmers(18) (1960). The ions leaving a
discharging point cause a reduction in the potential gradient downwind.
With one field mill upwind and another downwind the measured change in
potential gradient can be used to find the point current. Although the
method is indirect, no modification of the discharging object is necessary.
The authors found evidence that a tree in full leaf gives much less point
discharge current then heti previously been unstated, a matter of importance
in discussing the total charge brought to earth.
Mauer tanti Chttlmers(" (190) report mennurmeentu of potential gradient
and discharge current rrom a point fixed 2 m above a horse-chestnut tree
13 m high. For a given value of upwind speed their results shwa linear
relation between point current and potential gradient. These authors also
describe a new method of measuring point discharge current down a tree.
They drilled two holes through the bark of a lime tree, one .3 a above the
other, and inserted tubes containing mercury to nuke electrical contact with
the sapwood, vonnecting the leads to u galvanometer. This effectively short-
circuited that section of the tree and the current measured was almost all the
point diseharge Aurrent. Here too they round it linear relationship with
potential gradient, end utme indication that a tree given lens point current
when in leaf. Further observations with the same apparatus, by Chalmers (1962),
Underline the need to exercise caution when interpreting point discharge records.
He reports that the tree current is not only always less than that through an
artificial point, but that during the rapid field changes accompanying lightning
the tree current has a quite different course from that of the artificial point,
which follows the egrtmeter record in the usual way. An approximate relatien-
ship embracing the linear law for current to an earthed point has been deduced
theoretically by Chalmers(2) (1961); the current i 2xeVW where e In the
electric permittivity or air, V the potential gradient and if the wind speed.
(18) Durham University, England
(19) Durham University, England
(20) Durham University, England
. .
I.
��'� :0 ;
;.: kr:.�
$4.
.e
*!,44
d for Release. 2017/09/11 C06461858
� � Approved for Release: 2017/09/11 C06461858
10.
Using Adamson's field-change compensated exposed collector system (1960)
21) Ramsay and Chalmers f( 11960, have measured the current brought to earth
daring continuous Lon-stormy precipitation. The comparatively short time
constant of 20 zee enabled them to examine in greater detail than before the
connexion between current density and potential gradient. This was reasonably
linear for observations in the winter 195741 and or the well-known Vann
/ a a(F + C) where F is the potential gradient and a and C are constants.
Correlation was poor in the summer of 1958. The connexion in moat nearly
linear during sleet and "wet snow", and supports the earlierconclusions of
Chalmers (1956, 1959) that in nimbostratus clouds the precipitation, when
in the form of snow, receives a negative charge, leaving positive behind in
the cloud: but when it melts it acquires a positive charge, leaving negative
behind.
5. Thunderstorm Electricity
A new theory of thunderstorm electrification has been advanced by
Latham and Mason(22) (1961b). It is based on the results of their detailed
laboratory experiments which are in excellent agreement with their theory of
electric charge transfer associated with temperature gradients in ice by ft
kind of thermoelectric effect (1961a). To quote one of the authors, Mau=
(1961): "The positive hydrogen ions (protons) and the negative hydroxyl
ions (08-), fmmed by the thermal dissociation of a small fraction of the ice
molecules, bees= separated under the influence of a temperature gradient.
It we imagine a steady temperature difference maintained across a piece of
lee, the warmer end will initially pousess higher concentrations of both
positive and negative ions. Ions of both types will diffuse down this
concentration gradient towards the colder end, but because the mobility of
the positive ions is at least ten times that of the negative ions, they will
move ahead and produce an excess of positive charge in the colder part of
the ice."
This charge transfer process is considered to operate' in a storm when
supercooled water droplets captured by falling soft hail pellets, freeze on
(21) Durham University, England.
(22) Imperial College, London.
d for Release. 2017/09/11 C06461858
' � �
t� ;.;�������1
�
.�" ��� �
r -IY
! ����� .
,
oprov
Approved for Release: 2017/09/11 006461858
11.
contact, throw out small positively charged splinters, and so leave a
negative charge on the hailstone. This charged splinter production was
verified in the laboratory by the authors, working with a simulated hail
pellet growing by accretion of supercooled water droplets. They confirmed
the earlier experimental findings of Mason andlasybank(23) (1960) who
observed the splintering =clumsily negative charging, on freezing, or a
supercooled water droplet suspended on an insulated fibre. While a drop-
let is freezing it will have a liquid Centre at 00C and a solid outside
part at a lower temperature, giving a radial temperature gradient in the
ice shell. According to the charge transfer process there will be nn
excess of positive space charge in the outer layers of lee, and, when the
droplet bursts, the outer layer will tend to carry off positive charge,
leaving the residue negatively charged. Such negatively charged hailstones
=falling away from the positive splinters would produce a positive
dipole in agreement with that in a thundercloud. .
LeXham and Mason (1961b) proceed to calculate the rate of charge
production in a thundercloud. For soft hail pellets of average radius
and fall velocity v the volume swept out per aec Ica* v and so each pellet
makes saPne collisions with oupercouled&oplets in concentration ndLiM
is the collision efficiency. If there are nh halip�llets per unit volume
there are thus Ealndniiv collisions per unit volume per second. The soft
hail has an equivalent precipitation intensity p, i.e. the mass of water
4
falling per unit area per second given by rniOnhv where ; is the mean
density of the hall. In terms of p the number of collisions becomes
4E -AL- n � It the charge produced per droplet is q,, the rate or charge
production per unit volume is given by
dt d
Using the values E 1, p a5 cal b-1 or 5/3600 cm 0-4, 1-1 a 0.2 cm,
0.5 g cm-S, rid a 1 cm-S and the authors' laboratory value
qda x 10-43 e. s. u., we have
(23) Imperial College, London
pproved for Release: 2017/09/11 006461858
e .
�
� �
� qr.
5i-� � V'
�;,
:
��t-i.;'1 4 ;
�7
..,r.
�.:
:
�
�
:.�
Approved for Release: 2017/09/11 C06461858
12.
= 4 x 10-e e.e.u. ettleu-1
dt
tn 1 C ite-Smin-1
The authors consider this rate adequate to provide enough charge to give the
first lightning flash within about 20 minutes of the detection of precipitation.
particltsby radar, and they euggest that their theory gives the principal
mechanism of thunderstorm eleetrifIcutlon.
Latham andletson (leau,b) huve also investigated the charge produced
by the momentary contact of two pieces of ice at different temperatures.
From the temperature gradient charge transfer theory they predict a maximum
charge transfer of 3 x 10 AT e.s.u. enra, Where AT is the temperature-
difference, for a contact time of 0.01 seconds. For longer times the samples
rapidly cane to the sumo temperature end the charge will tend to zero. These
results were confirmed experimentally by the authors. Calculations of the
rate of charge production in a storm by this process, with hall:Mown; falling
through u cloud if lee crystals, give only 10-4 C itm-6 min-1, end the uuthure
conclude that although the sign of the charge on the httiletone will be
negative, as required, the contribution to stone electrification will be
only slight.
The theory of the charging of hail pellets by these two processes has
been extended by Latham and Mason (1962) to the case of collisions occurring
in polarizing electric fields of up to about 1000 V cd-1 as found in thunder-
storms. They also examined this question by laboratory experiment. They
conclude that such fields have little effect on the rate of charging predicted
by the main theory outlined above.
There lu a serious dlsercpuney between the unnerved charge for Ice-Ice
contacts reported ter laUnwi nnd Menton and that by Reynold, Drool( and
Gourley (1957), the latter being some five orders of magnitude larger.
There seem to be no other measured values, but Hutchinson(24) (1960) reported
that for momentary contact between two ice crystals grown frau the vapour and
having temperature differences up to lb deg C any charges due to the contact
(24) Durham University, Englund
Aso�
rt:
pproved for Release' 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
13.
were below the apparatus sensitivity of 6 x 10-6 e.n.u. The urea in contact
lay between 0.2 and 2 nce and the contact time between 0.2 and 0.5 sec. A
(25)
similar conclusion was reached by Evans (1962). Charges as large as
those reported by Reynolds, Brook and Gourley should have been detected
easily. The discrepancy has already led to some discussion by Reynolds
and Brook (1962) and Mason and Latham (1962).
Evans (1962) also has measured the charge romining when a supercooled
drop on a fibre freezes, bursts, andejects fruomeas. Although his results
refer to only 50 drops there is an indication that the charge produced is
often larger than the Latham-Mason theory can easily explain.
The production of ice splinters on exposing a frost deposit to an
airstream at different temperatures has been examined by Latham(26) (1962).
The particles were found to carry a charge, its sign aid magnitude depending
on the difference in temperature between deposit and airstream, and explained
by the Latham-Mason temperature-gradient charge transfer theory.
the
At/Electrical Research Association Laboratories at Leatherhead, Eugland,
an inexpensive and reliable lightning flash counter has been developed (Guide,
1962). It operates on positive potential gradient changes caused by negative
strokes to earth up to a distance of ho km. The recovery time is of the order
of 1 second so that if multiple strokes occur only one will be recorded.
Since the instrument is also triggered by the appropriate cloud to cloud
discharges it is necessary to know the ratios of negative to positive earth and
cloud strokes respectively.
(25) Durham University, England
(26) Manchester College at Seience and Tbchnulogy, England.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
34.
6. Referencen
Adamson, J., 1960, Quart, J. R. Met. Soc. 86, 252.
Adkins, Ca., 1959, Quart. J.R. Met. Soc. 85, 237.
Browning, K.A. and Ludlam, P.R., 1962, Quart. J.R. Met. Soc. 88, 117.
Chalmers, J.A., 1956, J. Atmos. Tem Phys. 311.
Chalmers, JAI., 1959, Recent Advances in Atmospheric Electricity,
p. 309, Pergamon Press, New York.
Chalmers, J.A., 1961, J. Atmos. Tem Phys. 24, 339.
Chalmers, J.A., 1962, To be published in J. Atmos. Ter. Phys.
Currie, D.R. and Kreielsheimer, K.S., 1960, J. Atmos. Terr. Phys. 22, 126.
Evans, D.G. 1962, Mc. Thesis, Univ. of Durham, England.
Flanagan, V.P.V. and O'Connor, T.C., 1961, Geofis. Pur. Appl. :2, 148.
Golde, R.B. 1962, Private communication.
Hutchinson, W.C.A., 1960, Quart. J.R. Met. Soc. 86, 406.
Keefe, D. and Nolan, P.J., 1961, Geofis, Pur. Appl. 22, 155.
Keefe, D. and Nolan, P.J., 1962, Proc. Roy. Irish Acad. 62, 43.
Keefe, D., Nolan, P.J. and Rich, T.A., 1959, Proc. Roy. Irish Acad., 60, 27.
Latham, J. and Mason, &J., 1961a Proc. Roy. Soc. A, 350, 523.
Latham, J. and Mason, B.J., 19616 Proc. Roy. Soc. 16 260, 537.
Latham, J. and Mason, B.J., 1962, Proc. Roy. Soc. A4 266, 387.
Latham, J� 1962, not yet published.
Lew, J., 1961a, Geofis. Par. Appl. 22, 53.
Law, J., 19616, Geofis. Put. W. 22, 102.
Mason, B.J., 1961, The Physics of Cloud, Rain and Lightning. P. 77,
Inaugural Lecture, Imperial College, London.
Met.
Mason, B.J. and Latham, J., 1962, Quart. J. Roy. Soc. 88, 551.
ihisono D.J. and Marbank. J., 1960, Quart. .7. Roy. Met. Soc. 86, .176.
Maund, J.E. and Chalmers, J.A. 1960, Quart. J. Boy. Met. Soc. 86, 85.
Milner, J.W. and Chalmers, J.A., 1961, Quart. J. Roy. Met. Soc., 1n5 592.
Nolan, 1955, Proc. Conf. Atmos. Elect. Geophysical Research Paper
No. 42, p. 113.
Nolan, P.J. and Kaman, E.L., 1949, Proc. RR,. Irish Aced, 23, 171.
O'Connor, T.C. ami Sharkey, W.P., 1960, Proc. Roy. Irish Acid. 61, 15.
, � ' �
pproved for Release: 2017/09/11 C06461858.
:�,;!:"�
Approved for Release: 2017/09/11 C06461858
11.
Pollak, I.. W. nud Metracks, A.14 1962, G.or1r. Pur. Appl. 51, 225.
Ramsay, M.W. and Chalmers, J.A., 1960, Quart. J. Roy. Met. Soc. 86, 530.
Reynolds, S. E. and Brook, M., 1962, Quart. J. Roy. Met. Soc., 88, 550.
Reynolds, S.E., Brook, M. and Gourley, M.F., 1957, J. Met. 14, 426.
Rich, MA., Pollak, L.W. and Metnieks, A.L. 1962, Geofis. Pur. Appl. 217.
Smiddy, H. and Chalmers, J.A., 1959, Quart. J. Roy. MOt. Soc. 86, 79.
Stein, J.M., 1958, M.Sc. Thesis, Univ. or Auckland, New Zealand.
Wildman, P.J., 1962, Ph.D. Thesis, linty. or Durluun, England.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
SESSION 7.3
ELECTROMAGNETIC ENERGY RADIATED FROM LIGHTNING
Atsushi KIMPARA
The Research Institute of Atmospherics
Nagoya University
Toyokawa
Japan
Abstract�This paper is to survey the study recently developed
on the electromagnetic energy radiated from lightning, i.e.
atmospherics, not including propagation. Characteristics of
the electrostatic, induction and radiation fields of lightning
are fully described, including the frequency spectrum in the
neighbourhood of the source. Consequently this paper will
supply a foundation to the study of propagation of atmospherics,
slow tail, ELF and VLF propagations, whistlers, swichanism
of lightning discharge, etc.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
I. Introduction
filattAVUL
This paper is to survey the generalof the develop-
ments of observdtion and theory which have been made recently
in the field of electromagnetic radiation from lightning and
at the same time to suggest the items of collaborated study
for the future.
In order to study the characteristics of lightning
discharge many kindSof measurement have been made and developed,
i.e. optical, photoelectric, electrostatic and electromagnetic
methods have prevailed all over the world. Here in this paper,
specifically, the characteristics of electromegnetic energy,
i.e. atmospherics in a broad sense, radiated from lightning,
not including propagation, are described.
The atmospherics proplwate through the space between
the ionosphere and the earth in the wave guide transimission
mode or in the ray mode reflecting between them. Some of
the energy penetrate the ionosphere into exosphere along the
geomagnetic line or force, and go to the other hepdaphere
where they are reflected back end return to the source again
along the same
geomagnetic line .4* force. As the exosphere
is the medium plasma with magnetic field, It is dispersive
A
and during the lourney atmospheric pulses become whistlers
from which the density of electro:, in the exosphere is
evaluated and the existence of proton in it is proved.
Frequency spectrum of atm,snherics at the various distances
from the source will show the propagetion characteristics
of LF and VLF waves. Since the long waves are not disturbed
by geomagnetic storms and proparate with low attenuation,
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
- � -
they are very nar"n1 to the international e�alprla,a1 .4. the
fre.n:ency standards an well as 10 the r:ethod (i. havl-ati,,n.
lt Is because the radio en.-Ineers and (.,eo..hysig's :Ink( mch
of the stuiy atmospheric and whistlers.
Consequently the invt.:.tiaation of characttristics of
electric fields in the neighlorhood or 11.-htnin.- discharze, nre
very useful to the study or the m.,chanism or 11,,htning nischaree,
the pro:nation of longer radio waves and the int.rfc.r(nce
of atmospherics to radio communications.
"or this purpose workrs have made so tigr the wPveform
measurement with wide hand rect-ivers, in which the hand
.,Idth is less thnn I. kc/s to avoid interffrence in ra!rly
clouded Wray frequency rrOon. ELF hand, 1 c/s-3000 c/s,
tlihh at recently attentions or anollneern and nclentintn,
!s measured with receivers of pass band I. c/s-90 c/s for a lower
frequency reaion and with waveform recorders for a highcr
frequency rein, "slow tall". For HP, VHF and UHF revions
ohaervations are :,.ado wit!. s!nzle frequency rreelvern or
42.
very narrow band width to avoid the int,:rfpcnce of radio
A
communications.
Lightning dischariTs tie, divided Into 2 clannes, 1. e.
the cloud-earth discharge and the intra-cloud discharr.e.
The cloud-earth dischnrae consists of the 'Ira�preliminary
eq.
d'sehar,7o, the prelirdnary ( Ti', !fruition stareAb,
hren n t a ta. , I , I it! 4* i ti r� tm,�.� find I �I r !kr I.. r
:;t4t;��� . , bp Hp i'MINI�1� r C hi! :1:1�:11 'iii
streamer ettwe, s, ?, the rinal dischar,-0 stay.,e, etc.
Corresponding to each of these optically observed stages,
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
it-
-3-
a characteristics change of electric field is observed with
(1)(2)(3)04)(5)(6)
fairy good response.
The intre-cloud discharge displays also characteristic
field changes somewhat different from those of cloud-earth
discharges. Therefore the investigation of electrostatic,
induction and radiation field of these stages and the
comparison among them are very useful to investigate the
details of characteristics of lightning phenomena. It is
because the electronical methods, developed remarkebly in the
last decade, are recommended to reveal quantitatively tLe
details or tho phenomena better than the optical methods.
II. Electrostatic Field
In accordance with the observation at distances 25-250
Ian, P1erce(3)found with capillary electrometers the relation
between the positive and negative slow mean field changes
with distance. The field obeys inverse cube relation and
corresponds to a change of electric moment of 110 coul-km,
I. e. to a field-change of 1 v/m at 100 km. It is well known
that near a storm most field-discharges are positive, while
as the distance of the activity increases negative field-
changes become more frequent.
For any particular yeur and for magnitudes less than
about 100 v/m, I.1e rutio Nt/N-, where Nf and M- are the
number of all positive and negative field changes, Is conste.A.
This constant value may differ from year to year, but there
is no significant change with magnitude between 100 and 0.1 v/m.
Above 100 v/m positive field-changes become increasingly
predominant as the magnitude of the field-change rises.
1.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-h-
A field-change of 100 v/m correseonsponds to e distance of
stout 20 km; the constancy of NO1- belowe 100 v/m therefore
implies that discharges, producing a reversal in the sign of
the associated field-change beyond 20 km, do not occur. The
changes from year to year in Nt/N-, for field changes, . tir � . ,e1h.
4 1 1 � ,
1
22 �
0 -1
"
s � gm. �
��.!kri�*,
-ot
- � Iiirlr�4006
. APAre_,
1.FqV
I �4;;
leP'���Tt
Fig.
2.
0 05
_L._ L__ _L___L 1 1 1 I S J L. .-J
10 1.5sec.
Comparatively irregular variation of frequency,
pproved for Release: 2017/09/11
not correlated.
Approved for Release: 2017/09/11 C06461858
Fig. 33. Multiple step variational types, not correlated.
10 15sec Fig. 35. Step variational
L ' I type, not correlated.
_ L L_. 1 1 __I. _1_ .1 1.
Whistler of usual type' urecoded by one riser,
pproved for Release: 2017/09/11 C06461858
.24 .
Approved for Release: 2017/09/11 C06461858
kc/s
8-. -rii.., r i 7
[
-fit..." � ' ..a; . q
itiz" � -, 4.,
6
�
4
"eaas.0
a.afkotAFI .a 44 -
-4:41fer* �
�A-
41,44-A r
/14.--75s ,r4 �
0
1
05
1.0
111111
VP;
I 15
2 Osec.
1-/
Fig. 37. Whistler of usual type preceded by two risers, correlated.
a-at
13. Special remarks of variational forms of unusual musical atmospherics.
The musical atmospherics of urusual types illustrated by Figs. 22-37
show plainly the most shifting variational aspects and forms. The conclu-
sion must be that within the thunderstorm atmosphere there exist some ve-
ry complicated, special modes of generation of unusual musical atmosphe-
rics. A theoretical explanation of the phenomena will for this reason re-
quire a more extended experimental analysis than has been obtained by the
results presented here. It can only be confirmed that a theoretical treat-
ment of the phenomena will be considerably more intricate compared with
what was valid for musical atmospherics of the usual type.
It is especially striking that the analysed unusual musical atmos-
pherics with or without correlation result in a conspicuous conformity
in their characteristic variational forme. This phow e that the unusual
musical atmoapherics originate from the same typo of sources within the
thunderstorm atmosphere. The conclusion is near at hand that the non-
correlated musical atmospherics are also caused by lightning discharges.
The problem that non-correlated musical atmospherics have also been
caused by liahtning discharges without having been recorded as correlat-
ed has its explanation in the limitations of the equipment used with re-
gard to distance sensitivity. For determination and recording of wave-
forms have been used CRO recorders and for the determination of the dis-
tances CHO direction finders. The acceptable distances for an accurate
determination were estimated to be 2000 km at best. It seems quite ac-
74,
aaa.7 ceptable that musical atmospherics beyond that distance have sufficient
propagation to reach the observation station and to be observed and re-
corded there as not correlated.
Another explanation of the existence of non-correlate:! musical at-
pproved for Release: 2017/09/11 C06461858
'J�_IL
Approved for Release: 2017/09/11 C06461858
25
14,
mospherics within the sensitivity limit of the station is possible. In a
thunderstorm region lightning discharges might occur producing musical
atmospherics characterized by effective propagation qualities. On the
other hand, in spite of their location within the sensitivity limit of
the station, lightning discharges have sometimes insufficient propaga-
tion effectivity to reach the station. This can very well explain why in
special situations musical atmospherics have not been correlated.
.2uz_L_Jata
A station for analysis of relations between lightning discharges
and musical atmospherics of usual (whistler) and unusual variational
forma has been operated for some years near Uppsala.
Recording cathode-ray oscillographs were used for the analysis of
the lightning discharges whose relations to musical atmospherics were
investigated. Cathode-ray oscillographic direction finders placed at two
stations with suitable distances between them made it possible to deter-
mine the sources of the lightning discharges investigated.
Through comparative harmonic analyses it was shown that lightning
discharges producing musical atmospherics cf the usual type - whistlers -
were characterized by a preponderance of frequencies around 5-S kc. Mul-
tiple lightning discharges were found to be followed by multiple whist-
lers.
The recording method of the station allowed also of an investigation
of correlations between lightning discharges and musical atmospherics of
unusual and irregular variational forms. It was found that out of 700
unusual musical atmospherics about 70 % were correlated to lightning dis-
charges and about 30 % were not. The striking variational resemblance
between correlated and non-correlated short-time variational types of un-
usual musical atmospherics indicated that the non-correlated variational
types must also emanate from lightning discharges.
Acknowledgements
Special cathode-ray recording oscillographs and direction finders
were used in this investigation. The instruments were constructed with
the help of grants from Statens naturvetenskapliga forskningsrAd (the
Swedish Natural Science Research Council) and Statens tekniska forsknings-
rAd (the State Council of Technical Research). Other instruments and
other experimental er_uipment used for this whistler investigation were
constructed with the support of a special grant from Statens naturveten-
approved for Release: 2017/09/11 CO6461858
Approved for Release: 2017/09/11
1-11:
2.6
skapliga forskningsrid.
G
The Swedish Board of Telecommunications has given valuable assist-
ance with telephones between two direction-finder stations and with the
construction of a special relay system for distance operation by tele-
phone of the whistler station.
� At the University of Lund the Institute of Physics and the Institute
ri
of Genetics made it possible to operate a direction-finder station used
IP
in the investigations.
For analysis of whistlers the author was permitted to make use of a
sound-spectrograph which belonged to the Institute of Phonetics at Upp-
sala University.
The author is very much indebted for the valuable help given.
The geophysics research reported in this article has been sponsored
and supported by Contract AF 61(052)-07, Geophysics Research Directorate
and also by Contract AF 61(052)-171, Electronics Division, Cambridge Re-
search Center, Office of Aerospace Research, United States Air Force
through its European Office, Brussels.
I, �
?'
17,
ADDrov d for Release. 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
27
;7-
References
1. BARKHAUSEN, H., Pfeiftone aus der Erde. Physik. Zeitschr. XX. 1919.
2. BARKHAUSEN, H., Whistling Tones from the Earth. Proc. Inst. Radio
Eng. 18, 7 (1930).
3. ECYERSLEY, T.L., A Note on Musical Atmospheric Disturbances. Phil.
Eag. 6, 49 (1925).
4. ECKERSLEY, T.L. and CHAPMAN, S., Radio Echoes and Magnetic Storms.
Nature 122, 768 (192F4).
5.
6. BURTON, E.T. and BOARDMAN, E.M., Audio-Frequency Atmospherics.
Proc. Inst. Radio Eng. 21. 1476-94, 10 (1933)-
ECKERSLEY, T.L., Musical Atmospherics. Nature 135, 104 (1935).
7. STOREY, L.R.O., An Investigation of Whistling Atmospherics. Phil.
Trans. Roy. Soc. B. 246, 113 (1953).
E., Wave-forms of Electric Field in Atmospherics Recorded
Simultaneously by Two Distant Stations. Arkiv Pit. Geo-
fysik 2, 10 (1954).
H., Magnetic field variations from lightning strokes in
vicinity of thunder storms. Arkiv f�r Geofysik 2, 20 (1956).
H. and KNUDSEN, E., Combined Analysis of baylight .Photo-
graphs of Lightning Paths and Simultaneous Oscillographic
Records. Recent Advances in Atmospheric Electricity. Book.
Pergamon Press, London 1959.
E. and VOLLYER, B., Variation Forms and Time Sequence of
Multiple Lirhtning Strokes. Arkiv for Geofysik 2, 25 (1957).
H. and KNUDSEN, E., The Relation Between Lightning Dis-
charges and Whistlers. Planetary and Space Science, Perga-
mon Press 1959. Vol. 1, pp 173-1q3. Printed in Great
Britain.
9.
NORINDER,
q. NORINDER,
10. NORINDER,
11. NORINDER,
12. NORINDER,
13. NORINDER,
14. NORINDER,
15. NORINDER,
H. and ENUDSEN, E., Lightning discharges as a source of
whistlers. Arkiv for Geofysik 3, 11 (19(0).
H. and FNUrSEN, E., Recent aesults in the Investigation of
the helation between Lightning Discharges and Whistlers.
Planetary and Space Science, Pergamon Press 1961. Vol. 5,
pp 46-49. Printed in Great Britain.
H. and KNUDSEN, E., The dispersion of whistlers compared
with the geomagnetic latitudes of their sources. (Research
note). Planetary and Space Science, Pergumon Press 1961.
Vol. 5, pp 326-32S. Printed in Great Britain.
16. NORINDER, H. and KNUDSEN, E., Multiple lightning discharges followed
by whistlers. Arkiv for Geofysik 3, 12 (1960).
17. NORINDER, H. and IMUDSEN, E., Occurrence of different kinds of
whistler activity. Arkiv for Geofysik 3, 17 (1961).
E. and NORINDER, II., Different types cf music-,i1 atmospherics
and their relations to lightning discharges. Arkiv for Geo-
fysik 4, 5 (1962).
W.C., The Current-Jet Hypothesis of Whistler Generation.
Journal of Geophysical Research 65, 7 (1960).
19. KNUDSEN,
19. HOFFMAN,
t
I
pproved for Release: 2017/09/11 C06461858
N.!
Approved for Release: 2017/09/11 C06461858
20
20. WAIT, J., Symp. on Prop. of V.L.F. Radio Waves, Boulder, Colo. 4,
23 (1957).
21. WAIT, J., The Attenuation vs Frequency Characteristics of V.L.F.
Radio Waves. Proc. Inst. Radio Eng. 45, 768 (1957).
22. NORINDER, H., KNUDSEN, E. and VOLLMER, B., Multiple Strokes ir
Lightning Channels. Recent Advances in Atmospheric Electri�
city. Book. Pergamon Press, London 1q58.
23. NORINDER, H. and KNUDSEN, E., Lightning discharge paths photographed
in daylieht and analysed by simultaneous variations of
magnetic field components. Arkiv for G.eofysik 3, 19 (1961).
;
-!!'�
go-
. ..
1
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Problems of Fair Weather Electricity;. Introducing Remarks.
It is a task of our conference to review our knowledge and to recom-
mend further investigations on the different branches of atmospheric
. When I try to do this for the field of the "Fair Weather Electricity"
("FWE") I have the feeling that some of you may think:
Why still FE? This is overdone t Moreover the results
in this field are so complex and so contradictory that
it seems senseless to continue!
But as scientists, I believe we should hesitate to use such an argu-
ment to resign in view of difficulties. Besides there are quite dif-.
ferent opinions concerning this question. Therefore let us examine
phenomena connected with disturbed weather. Today rather the oppo-
is true as we may see, e.g., when we look on the program of the pres-
The division of the atmospheric electricity in the two parts; namely,
the FWE and the "Disturbed-Weather Electricity," was justified afterwards
by the dynamic conception of the atmospheric electricity: We have to
distinguish generation - that is the disturbed weather electricity, or
more precisely, the thunderstorm electricity, where the charges are separ-
ated - and consumption - even the FWE, where the separated charges will be
d for Release. 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
neutralized. So the two parts are connected together very closely, having
the same importance in the explanation of the whole picture.
If we look now on the fair weather electricity we first remember the
�fundamental discoveries of the last thirty years: We understand why we
find everywhere and always an electrical field in the atmosphere charac-
terized by typical periodical and nonperiodical variations. But these
results emerge only if they are based on a large qeantity of data, while
fcr shorter series of observations the average picture is disturbed more
and more by local conditions and meteorological influences. In other words:
we got a climatological picture only, based on statistical evaluation.
However, this method is quite insufficient to relate, for instance,
the atmospheric electric with the meteorological phenomena. Thus, in my
opinion investigation in this branch is rather underdone than overdone.
LL. Modern Problems in FWE
A. The Stratosphere
. Let us lock now on the modern problems in FWE. As you know, it is
one of the most difficult problems in this branch that frequently the
results are ambiguous because, in general, there are worldwide influences
superimposed on local effects - especially if we evaluate measurements
near the ground. Therefore, we have to look first for suitable methods
to separate the two districts of influences.
Thts may be achieved when we try to separate the researches with
respect to space; We have to distinguish at least two spheres of a quite
different behavior, the troposphere and the stratosphere/mesosphere.
Furthermore, we usually divide the troposphere in two regions, a lower
one which is characterized by the vertical turbulent connection, and an
upper one which is governed generally by a horizontal movement of the air.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
- 3 -
It is evident that the processes and the problems we have to clear
up are quite different in these three regions, from the meteorological
point of view as well as from the electrical one:
The elctric behavior of the stratosphere/mesosphere is governed
essentially by worldwide steering effects of the main generator and by
'varial-,lons of the ionization according to the latitude effect of the
cosmic radintion only. On the other hand, tropospheric variations of the
atmspheric electrical behavior are controlled by the influence of the
naustauschn on the aerosol conditions.
Thus, I believe we can separate the "targets" of further researches
in FWE in the two general groups of
stratospheric researches of the electric field and
the conductivity and of
(2) tropospheric results of the aerosol conditions.
Let me give some examples:
If we look on the first group, i.e. the stratospheric problems, we
need first systematic measurements of the total potential difference V
between the surface of the earth and the ionosphere and of their variations
in time. Results of this kind can be obtained only by measurements in
higher altitudes and at places not or less effected by the "austausch"!
So we have there a special task for the aerological method of measuremerit
as developed In the last time.
Two ways seem to be successful:
(a) Integration of field measurements by airplanes or radio-
sondes (0.H. Gish, 1944; J.F. Clark, 1956; J.H. Kraakevik,
1958; H.J. Fischer, 1962 et al.)
(b) Direct measurements of V and dV/dt in about 11 km altitude
by airplanes or constant level balloons with radiosondes -
pproved for Release: 2017/09/11 C06461858
' Approved for Release: 2017/09/11 C06461858
as proposed by H.W. Kasemir, 1950.
Measurements of this kind promise a better insight into the mechanism
of the worldwide atmospheric electrical. circuit. I think there is no doubt
that our basic hypothesis is correct, but up to now it is based only on
mean results of a few polar expeditions and on those of the cruises of the
Carnegie Institution. We are not able to control this in detail, e.g. to
find out changes of the thunderstorm activity in different parts of the
world - a target of special interest for the meteorology as well as for
aviation.
There are only first hints in this direction (H. 'Bran and
E. Theunissen, 1957 H. Isran, 1957).
The aerological researches may be aided by continuous records oc the
atmospheric elements over the free oceans and at mountain tons. The first
one could be done on the "weather ships" fixed now at different places of
the oceans. Furthermore, I believe we should ask the permanent stations
in the arctic and antarctic regions for atmospheric electrical records.
Another urgent question concerns the conductivity in the stratosphere/
mesosphere. Here we have only scarce results up to now. The conductivity
was measured directly, e.g. by radiosendes, up to about 30 km (C.G. Stergic.,
S.C. Coroniti, A. Nazarek, D.E. Kotas, D.W. Seymour, and ,L1/. Werme, a955
R.H. Woessner, W.E. Cobb and R. Gunn, 1958). Furthermore the behavior of
electromagnetic waves in the higher atmosphere allows to measure the
density of electrons or the conductivity respectively in the ionosphere,
i.e. for altitudes above 80 km. In the interval between 30 and 80 km the
conductivity values were computed (H. Israel and H.W. Kasemir, 1949;
R.E. Bourdeau, E.C. Whipple, Jr. and. J.F. Clark, 1959) and measured directly
only once by a rocket sonde (R.E. Bourdeau et al., 1959).
This first experiment shows already a considerable discrepancy in both
the computation and the measurement, especially for the region above 50 km
pproved for Release: 2017/09/11 C06461858
.-haight*
Approved for Release: 2017/09/11 C06461858
ThereforemeaSurements of this kind should be continued
;let Us, 109k Into .Some of. the most important problems In this
(a) In the region of 30 km to 70 km the electron concentration
is influenced by the intensity of ionizing radiation, by the
recombination processes, and by electron detachment and
attachment rates. To understand. which of these quantities
prevail in physical processes throughout this region, it is
necessary to determine the chemical, molecular, and at:mile
components and their density, including their change with
(b) In'lower altitudes the conductivity of the atmospneres is
1.11determined by the density of the small ions43.4-*Alie
;44
ionosphere the conductivity is caused 'practically by free'
The transition is to be expected in the region.2
between 30 and 80 km of height.
More knowledge of this kind will give 'a better understanding,-
of the lower boundary of the ionosphere.
It also will yield more information cn-the global current:
Circuit of atmospheric electricity, concerning questions
such as the height distribution of the equalizing current,
latitude effects, field gradients in horizontal directions,
perhaps daily variations, etc. .
(e) Finally we may find in this region connections to geomagnetic
events, solar influences, aurora and similar phenomena.
It is true, measurements of this kind may be much more difficult to
carry out than the usual atmospheric electrical measurements, however, I
believe the adaptation of measuring equipment to rockets and satellites is
a technical and not a principal problem0
proved for Release' 2017/09/11 C06461858
Approved for Release: 2017/09/11 006461858
� B. The Tropouhere
A quite different picture of the electrical behavior we meet in the
troposphere;. we find a group of problems of another kind.
To characterize the situation we may remember the opinion of Lord Kelvin
100 years ago - that in the future the forecast of the weather would be done
with the electrometer. This prediction, of course, was a too optimistical
one; however, the essential point which provoked that statement is the same
up to the time being: All meteorological events are accompanied by charac-
teristical changes of the electric parameters.
What we have to do is to explain these connections and to classify
the electric variations. Maybe this is easier to say than to do because at
the first sight the results up to now seem to give a chaotic picture. We.
remember, e.g., wide ranges of variation spectra of the different elements
including fluctuations from the annual variations down to the so-called
"noise," the different combinations of the "electrode effect" with the
aerosol conditions, the radioactive influences, the movements of air masses,
etc. Although we know many details in this field, it is hard to find an
integral view up to now.
This, I believe, is the reason why some people voiced the opinion it
would be senseless to continue researches of this kind. But in my opinion
we have here no more difficulties than in the field of meteorology in
general. Therefore, rather we should examine our measuring methods if they
are adequate for the problems arising here.
As mentioned above in the troposphere we have to distinguish two regions
of a quite different behavior, i.e. the "Exchange Layer" and the upper tro-
posphere. In addition to this we have to separate a third region near the
surface of the earth which is governed by the so-called "electrode effect."
pproved for Release: 2017/09/11 006461858
of,'0
Approved for Release: 2017/09/11 C06461858
- 7 -
The boundaries between these spheres are fluctuating according to the
specific weather conditions, to the time of day, and to that of the season.
1. The Electrode Effect
The electrode effect is caused by the electric field in ionized
air near the electrodes, i.e. here near the surface of the ground. Con-
sidering the atmospheric condtions one can compute that this effect will
be essential up to an altitude of one or at most a few meters. This alti-
tude is smaller if the content of condensation nuclei in the air is greater
(J. Scholz, 1931), and the effect is depending on the ionization conditions
near the ground. They may limit it sometimes to the first decimeters above
the ground (A.R. Hogg, 1935; J.A. Chalmers, 1946 et al.)
Summarizing the results we find only a rough conception of this
effect. Especially we miss researches of the meteorological influences.
Therefore, I believe we have to see here a first important problem for our
future researches.
How will the electrode effect be influenced by the
meteorological conditions, the radioactive conditions
in the ground and in the air near the ground, the
aerosol conditions etc?
Moreover the region of the electrode effect is accessible easily for all
measurements and recordings we need. This enables us to examine the
meteorologic-electrical connections in a small range, so to speak.
Of course, the usual measuring methods will be insufficient for
researches of this kind. We have to measure at least the electrical field
. .
strength and the conductivity in small regions. Therefore, all kinds of . .
disturbances should be avoided as much as possible as they occur due to the
orographic situation, the installation or the working method, the equipment,
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
- 8
etc. It(is true, this will raise the claims for our measurements. However,
I be1iev4 the problems we have to investigate in the region of the tropo-
sphere require urgently a new effort in our work.
2. The Exchange Layer
In this region of the troposphere we meet in general quite
similar problems as mentioned before. However, the researches are more
difficult because this sphere is much more extended in both the horizontal
and the vertical direction. Therefore, we have to combine measurements
both near the ground and in the free atmosphere.
Looking into these problems we have to find out above all the
numerous influences of the "austausch" on the aerosol conditions. They
will give us the key to understand most of the meteorological electrical
relations.
The next step will be a systematical examination of the effect
of air mass movements, considering their aerosol conditions, their content
or radon, thoron, and decay products, maybe the content of fission products,
etc.
Although the measuring methods are insufficient here, too, a lot
of results came out already. I recall your attention, e.g., (1) to the
explanation of the different types of diurnal variations of the electrical
elements (J.G. Brown, 1930, 1935; H. Israel, 1948, 1950, 1952); (2) to the
explanation of the so-called "sunrise effect" (H.W. Kasemir, 1956); (3) to
the researches of the so-called "brightness effect" (G. Fries and H. Dolezalek,
1956); (4) to the "noise" of atmospheric electrical elements (H. Israel, 1958,
1959); (5) to the steplike variations of the atmospheric electrical elements
at the upper boundary of the exchange layer (F. Rossmann, 1950; Callahan,
Coroniti, et al., 1951) and to other ones.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
- 9 -
However, these results yield with few exceptions, average values,
describing the climatological behavior. This may be the reason why we miss
a systematic picture of the connections between the weather and the atmos-
pheric electricity up to now. Therefore, we have to look for a suitable
extension of investigation. We shall see that here, too, the usual measur-
ing customs must be changed.
"Off-hand-solutions" and "ad-hoc-theories," as they are tried some-
times, do not help us. They fail today as they failed in the days of F. Exner*.
3. The Upper Troposphere
The problems arising here may represent the last step in the new
program for atmospheric electrical researches.
Since the aerosol content in the air above the exchange layer in
general is unimportant, (see, e.g. R.C. Sagalyn and G.A. Faucher, 1954, 1955)
we can expect to be confronted in this region, first of all, with the influ-
ences of air masses and their movements, with effects of variations of radio-
activity, and with stratospheric influences.
Researches in this region will be done with airplanes, gliders,
radiosondes, and constant level balloons. Furthermore, it will be very help-
ful to record the atmospheric electric parameters at mountain tops of suffi-
cient altitude.
First, results and important hints for future researches will be
found, e.g., in the papers of F. Rossmann (1950); R.C. Callahan et al, (1951);
R.C. Sagalyn et al, (1954, 1955); C.G. Stergis et al. (1955); L. Koenigsfeld
(1955, 1957, 1958); J.F. Clark (1956, 1958); J.H. Kraakevik (1958); K. Uchikawa
*) so, e.g., the hypothesis of F.M. Exner (1886/1890) concerning a transport
of charges by evaporation of water, which was refuted by H. Benndorf ( ** )
and P. Lenard (1944); a revival of this hypothesis by R. MUhleisen (1958)
was refuted by H. Israel and R. Kno (1962; see also R. Knopp 1961).
**) H. Benndorf conducted in 189771898 field investigations in Siberia, which
demonstrated that the mechanism as suggested by Exner is not verified.
ip
ow. 'Ai
pp, -'041T:
:f4
� �
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
- 10 -
(1961); G. RBnicke (1963); and other ones. For observations on mountain
tops see, e.g., the researches of R. Holzer et al. (1955) in California
and Hawaii and of H. Israbl et al. (1957) in the alps.
III. The Measuring Methods
The researches proposed above require changes and improvements of the
measuring methods.
First of all we have to look on the comparability of the results.
For this the following demands must be fulfilled
(1) It is known that the so-called "Reduction on the Free Plane"
involves a considerable uncertainty, because it is impossible
to include the influence of the space charges. Therefore the
necessity to reduce the observed values must be avoided. In
other words, all measurements near the ground - especially
those of the electric field - should be done on an open plane
of sufficient size. The rules of H. Benndorf (1900, 1906)
may be used for the critical examination what means "sufficient."
This should be applied also to measurements in mountain regions
where we have to look for planes of sufficient size (plateaus,
glaciers, etc.).
Of course by measurements with aircraft, radiosondes, etc., a
computation factor concerning the geometrical forms is inevit-
able.
(2) Measuring techniques which disturb the natural conditions should
be avoided as completely as possible. This concerns first of
all the use of radioactive collectors for measurements near the
ground. For airborne equipments the collector may be used pro-
vided that the aspiration is sufficient. When using radiosondes
it is important to consider the researches of G. Remicke (1962)
pproved for Release: 2017/09/11 C06461858
'Approved foi.Release26i7/09./11-e06461858
- 11 -
concerning the mutual influences of the two radioactive
collectors.
(3) Different measuring equipments both for measurements near the
ground and in the free atmosphere should be compared by simul-
taneous application at the same place and over a longer period
of time.
(4) All researches should include simultaneous measurements of
the three parameters of Ohm's law, i.e. the potential gradient
(field strength), the conductivity (conductivities), and the
air-earth current density.
To avoid misunderstandings concerning the "sign," the remarks
of H. Israel (1961, 1963) may be mentioned.
(5)
For synoptical researches as proposed by H. Isragl (1954, 1955, 1961)
and included in part II, B,2 of this paper the following difficulty should
be considered.
(6) Synoptical researches near the ground will be disturbed by
the electrode effect which may be different at different sta-
tions. In order to avoid this difficulty it was proposed by
H. Isragl (1962) to measure no more at the ground itself but
in an altitude of at least two meters. If this proposition
will be accepted the comparability may be improved. In this
connection I like to refer to the researches of W.D. Crozier
(1963). He tested a new method of field measurements which
works with a minimum of disturbances.
IV. Some Indications for Practical Applications
Someone may ask for practical applications, if he thinks of the proposals
given above for further investigation on Fair Weather Electricity, and the
pproved for Release: 2017/09/11 006461858
Approved for Release: 2017/09/11 C06461858
- 12 -
'expense connected with it. It is true, scientific work will not be criticized
from this point of view; but, I believe we can answer questions of this kind
also. Let me give some examples.
(1) At first I like to mention here the method of M. Kawano (1958)
to evaluate the naustauschn and its daily variation on the
basis of atmospheric electrical measurements. Similar researches
were done by W.B. Milin (1951, 1953, 1954). Researches of this
kind will be very helpful for both climatological and meteor-
ological purposes.
(2) Some results concerning the ionizing effect of artificial
radioactivity in the air (see e.g., D.L. Harris, 1955; E.T.
Pierce, 1959; G. Kondo, 1959; K.H. Stewart, 1960; A. Oster, 1963
et al.) suggest the application of atmospheric electrical obser-
vations for watching the fission product content in the air.
(3) Some years ago was discovered that the atmospheric electrical
elements ondergo specific variations about 1 to 2 hours before
the onset of fog and about 1/2 to 1-1/2 hours before the dissipa-
tion of fog (see e.g., H. Dolezalek, 1957; G.P. Serbu and E.M.
Trent, 1958; L.H. Ruhnke, 1961 et al.) The application of this
results to the forecast of fog and fog dissipation will be of
special importance for the practical meteorological work
(H. Dolezalek, 1962).
(4) Other possibilities for the application of atmospheric electri-
cal results to practical problems came out from the researches
of A. Gockel (1915) and others, concerning the prediction of
thunderstorms; the results of J. Scholz (1935), concerning the
prediction of blizzards; and the observations of G. Rott (1963)
pproved for Release: 2017/09/11 C06461858
�
Approved for Release: 2017/09/11 C06461858
- 13 -
concerning connections between the behavior of the electrode
effect and the weather development during the day. - In all
cases the prediction arised from observationn during Fair
Weather many hours before the event in question.
' 'proved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
- 14 -
-REFERENCES:
BENNMORF, H. 1900 : Dber Stbrungen des normalen atmosphRrischen Potential-
gefIlles durch Bodenerhebungen. Sitz.Ber.Akad.Wiss.
Wien 109, 923...940
1906 : Dber gewisse Sthrungen des Erdfeldes mit RIcksicht auf
die Praxis luftelektrischer Messungen. Sitz.Ber.Akad.Wiss.
Wien 115, 425...456
BOURDEAU, R.E., E.C. WHIPPLE, Jr., and J.F. Clark 1959 : Analytic and Exper-
imental Electric Conductivity- between. the Stratosphere
and the Ionosphere. journ.Geophys.Res. 64, 1363...1370
BROWN, J.G.
1930 : The relation of space-charge and potential gradient to
the diurnal system of convection in the lower atmosphere.
Terr.Magn.Atmos.Elect. 35) 1...15
1935 : The local variation of the earth electric field.
Terr.Magn. 40, 413...424
CHALMERS, J.A. 1946 : The Ionization of the Lowest Regions of the Atmosphere.
Quart.Journ.Roy.Met.Soc. 72, 199...205
CLARK, J.F. 1956 : The fair-weather atmospheric electric potential and
its gradient. Diss.U.Maryland 1956
CALLAHAN, B.C.; S.C. CORONITI; A.J. PARZIAIE; and R. PATEN 1951 : Electrical
Cmductivity of Air in the Troposphere. journ.Geophys.
Res. 56, 545...551
CROZIER, W.D. 1963 : in print: Journ.Geophys.Res. 68
DOLEZALEK, H. 1957 : Remarks on the electrical conditions during disturbed
weather. Contract AF61(514)-640, Techn.Note No. 12
: Examples for a. synoptic evaluation of tJie measuring
results of four atmospte.ric-electric stations in
F63.(51.4)-640, Techa.Note No. 16
Swit2t-rland. Co!strarA A
1962 : Atmospheric Electric Parameter Study: Survey on an
Effect Relating Atmospheric Electric Variatlons with
Formation and Dissipation of Fog. Contra2t Nonr
3388(00) Fin.Rep.
EXNER,F. 1886/1890 : lber die Ursache und die Gesetze der atmosphlrischen
ElektrizitRt. Sitz.Ber.Akad.Wiss.Wien (ITa) 93 (1886)
222...285; and: 95 (1887) 1084 ff; 96 (1887)-419 ff; 97
(1888) 277 ff; 98-(1889) 1004 if; 99 (1890) 601 if. --
FISCHER, H.J. 1962 : Die elektrische Spannung zwischen Ionosphare und Erde.
Diss .Stuttgart 1962
FRIES, G., and H. DOLEZALEK 1956: On parallelism occurring in the registration
of potential gradient and general local brightness.
Contract AF61(514)-640, Techn.Note No. 7
pproved for Release: 2017/09/11 C064618581.111111111111111111111P
GISH, O.H.
Approved for Release: 2017/09/11 CO6461858
- 15 -
1944 : Evaluation and interpretation of the columnar resis-
tance of the atmosphere. Terr.Magn. 49, 159...168
GOCKEL, A. 1915 Zur Gewittervorhersage. Das Wetter 32, 121...124
HARRIS, D.L. 1955 : Effects of radioactive debris from nuclear explosions
on the electrical conductivity of the lower atmos-
phere. Journ.Geophys.Res. 60, 45...52
HOGG, A.R. 1935 : Mem. of Comm., Solar Observ. Canberra 7
HOLZER, R.E. 1955 : Studies on the universal aspect of atmospheric elec-
tricity. Contract AF19(122)-254, Final Report
ISRAEL, H.
1948 : Zuni Tagesgang des luftelektriscben Potentialgefalles.
Meteor.Rundschau 1, 200...204
und H.W. Kasemir 1949 : In welcher Hebe gebt der liftelektrische
Ausgleich vor sich? Annales de Geophysique 5, 313...324.
1950 : Luftelektrisdhe Tagesgange und Luftkerper (Studien
Eber das atmospharische Potentialgefalle III). Journ.
Atmosph.Terr.Phys. 1, 26...31
1952 : The diurnal variation cf atmospheric electricity as
a meteorologico-aerological phenomenon. Journ.Meteor.
9, 328...332
1954 : Luftelektrische Synopsis. Mitt.Dtsch.Wetterdienst,
No. 7
1955 : Synoptic Researches on Atmospheric Electricity. Proc.
Conf.Atmos.Electr., AFCRC Geophys.Res.Rapera No. 42
11...20
1957 : Atmospheric Electric and Meteorological Investigations
in High Mountain Ranges. Contract AF61(514)-640,
Final Report
und E. THEUNISSEN 1957 : Luftelektrisches Potentialgefalle und
Weltgewittertatigkeit ein Beitrag zur grossraumigen
luftelektrischen Synopsis. Naturwissenschaften 44, 8
et al., 1957 : Atmospheric Electrical and Meteorological Investi-
gations in High Mountain Ranges; Microfilm of the
Results. Contract AF61(514)640, Appendix to the
Final Report.
1958 : The atmospheric electric agitation. In: Recent Advances
in Atmospheric Electricity, ed. by L.G. SMITH, Pergamon
press 1958, p. 149...160
1959 : The atmospheric electrical agitation. Quart.Journ.Roy.
Meteor.Soc. 85, 91...104
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-16-
1961 : Atmosphdrische ElektrizitRt, Tell II: Felder, Ladungen
StrBme. Leipzig,Akadem.Verlagsges.Geest & Portig KG
1962 : The comparison of atmospheric electric results. Journ.
Atm.Terr.Phys. 24, 65...66
1963 : The sign of the atmospheric electrical field. Archly
Meteor.Geophys.Bioklimat. (A), in print
und R. KNOPP 1962 : Das Prdbelm der Ladungsbildung beim Verdampfen.
Staub 22, p. 19 (see also R. KNOPP: Versuche zur
Ladungstrennung bei PhasenRnderungen, Verdampfung
und Kondensation. Thesis Techn.Hochsch.Aachen, Germany
1961).
KASEM1R, H.W. 1950 : Studien 'Aber das atmospharische Potentialfefglle IV:
Zur Strbmungstheorie des luftelektrischen Feldes I.
Archly Meteor.Geophys.Bioklimat (A) 3, 84...97
1956 Zur StrBmungstheorie des luftelektrischen Feldes III:
Der Austauschfenerator. Archly Meteor.Geophys.Bioklim.
(A) 9, 357.370
KAWANO, M. 1958 : The local anomaly of the diurnal variation of the
atmospheric electrical field. Res.Electrotechn.Lab.
(Japan) No. 569, July 1958
�
The local anomaly of the diurnal variation of the
atmospheric electrical field. In: Recent Advances in
Atmospheric Electricity. ed. by L.G. SMITH, Pergamon
Press 1958, 161...174
KOENIGSFELD, L. 1955 : Study of the variation of potential gradient with
altitude and correlated meteorological conditions.
Proc.Conf.Atm.Electr., AFCRC,Geophys.Res.Pap.No. 42,
21...25
1957 : La mesure du gradient du potential et de la conducti-
bilite par radiosonde. In: Atmospheric Electricity
During the IGY, ed. by H. ISRAEL, Aachen 1956, p.84...90
1958 : Observations on the relations between atmospheric
electrical potential gradient on the ground and in
altitude and artificial radioactivity. In: Recent
Advances in Atmo:Theric Electricity, ed. By L.G. SMITH,
Pergamon Press 1958, p. 101...109
KONDO, G. 1959 : The recent status of secular variations of the atmos-
pheric electric elements and their relation to the
nuclear explosions. Mem.Kakioka Magn.Observ. 9, 2...6
KEAAKEVIK, J.H. 1958 : Electrical conduction and convection currents in the
troposphere. In: Recent Advances in Atmospheric Elec-
tricity, ed. by L.G. SMITH, Pergamon Press 1958, 75...88
rt� '
pproved for Release: 2017/09/11
Approved for Release: 2017/09/11 C06461858
-17-
LENARD, P.
MILIN, W.B.
1944 : Probleme komplexer MolekEle. Sitz.Ber. Heidelberger
Akademie der Wissenschaften Va, Abhandlung. No. 27,
28, 29.
1951 : (New methods for the destination of the vorticity
coefficient in the atmospheric layer close to the
ground by using atmospheric electric factors -
RUSSIAN) USSR Glavnoye Upravleniye Godrometeor.
Slushby, Inform. Sbornik 1, 28...36
1954 : (Anomalous electric fields in the atmosphere -
RUSSIAN) Dokl.Akad.Nauk SSSR 95, 983...986
i S.G. MALAKHOV 1953 : (Air conductivity and turbulent mixing in the
1963
OaLEE, A.
PiEECE, E.T.
1959
RONIaKE, G.
1962
1963
ROSSMANN, F.
1950
RUHNKE, L.H.
1961
SAGALYN, R.C., and
SCHOLZ, J.
atmosphere - RUSSIAN) Izv.Akad.Nauk SSSR, Ser.
Geofiz., No. 3, 264...270
: in print: Atomkernenergie 8
: Some calculations on radioactive fallout with especial
references to the secular variations in potential
gradient at Eskdalemuir, Scotland. Geofisica pura
e applicata 42, 145...151
: Erfahrungen mit luftelektrischen Sondierungen in
der freien AtmosphRre. Zs.f.Geophysik 28, 105...126
: in preparation: results of atmospheric electric
radiosonde ascents in San Salvador 1957/58
: Luftelektrisches Feld und Wetter. Geofisica pura
e applicata, in print
: The Change of Small Ion Density Due to Condensation
Nuclei and the Relation to the Extinction Coefficient
of Light. Proc.Intern.Conf. on Ionization of the Air,
Philadelphia, V-1...V-13
G.A. FAUCHER 1954 : Aircraft investigation of the large
ion content and conductivity of the atmosphere and
their relation to meteorological factors. Journ.Atm.
Terr.Phys. 5, 253...272
1955 : Aircraft Investigation of the Large-Ion Content
and Conductivity of the Atmosphere. Proc.Conf.Atm.
Electr., AFCRC, Geophys.Res.Papers 42, 27...41
: Investigation of charged nuclei in the free atmosphere
Geofisica pura e applicata 31, 182
1931 : Theoretische Untersuchungen Eber die Feld- und Ion-
enverteilung in einem stromdurchflossenen Gas, das
auch schwer bewegliche Elektrizitatstrager enthalt.
Sitz.Ber.Akad.Wiss.Wien (ha) 140, 49...66
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
- 18
SCHOLZ, J. 1935 : Luftelektrische Messungen auf Franz Josephs-Land
(cont.) wahrend des Zweiten Internationalen Polarjahres
1932/33. Transact. of the Arctic Ipstit. USSR
16, 5...169
SERBU, G.P., and E.M. TRENT 1958 : A study of the use of atmospheric-
electric measurements in fog forecasting. Trans.
Am.Geophys. Union 39, 1034...1042
STERGIS, C.G.; S.C. CORONITI; A. NAZAREK; D.E. KOTAS; D.W. SEYMOUR; and J.V. WERME
1955 : Conductivity Measurements in the Stratosphere.
Proc.Conf.Atm.Electr., AFCRC, Geophys.Res.
Papers 42, 43...52
: Conductivity Measurements in the Stratosphere.
Journ.Atm.Terr.Phys. 6, 233...242
STEWART, K.H. 1960 : Some recent changes in atmospheric electricity and
their cause. Quart.Jcurn.Roy.Meteor.Soc. 86,
399...405
WOESSNER, R.H.; W.E. COBB; and R. GUNN 1958 : Simultaneous Measurement
of the positive and negative light-ion conductivities
to 26 kilometers. Journ.Geophys.Research 63, 171...180
UCHIKAWA, K. 1961 : Atmospheric Electric Phenomena in the Upper Air over
Japan; part I and II; Geophys.Mag. (Japan) 30, 619...672
pproved for Release: 2017/09/11 C06461858
I. RADIOACTIVITY AND THE INTENSITY OF IONIZATION 1
1. Intensity of Ionization 1
2. Intensity of Ionization Several Meters
Above the Ground 3
-.3. Intensity of Ionization in the Altitude 4
a) Radiation from the ground
b) Derivates of En & Tn in suspension 4
c) Lower Stratosphere 6
: 4. Artificial Radioactivity 7
. qualities of Small Ions 8
General Remarks - 8
Mobility � 8
Diffusion Coefficient 10
10
Recombination Coefficient 11
6. Small Radioactive Ions 12
II. AIR POLLUTION AND LARGE IONS 14
iii
2. 0:)tical Proi)(xtics 16
7. Condensation 17
4. Electric Charges 18
5. Mobility and Diffusion Coefficient 18
6. Coagulation 20
� 7. Radioactive Condensation Nuclei 21
8. Attachment of Small Ions to Charged and
1. General Remarks
Approved for Release: 2017/09/11 C06461858
Action of Radioactivity and of Pollution upon Parameters of Atmospheric
Electricity
Table of Contents:
Abstract
Page:
1
Mean Free Path
Neutral Nuclei 23
9. Attachment Coefficient Expression 24
a) Without Consideration of Small Ion's
Mean Free Path 25
b) Introduction of the Mean Free Path
of Small Ions 28 ,
pproved for Release: 2017/09/11 C06461858
'
' �
01.4.! �:..!
444'
4."J".4t/
.t!..t.:1/011111Nie
rre,
�
$�,..!-.1X7
Approved for Release: 2017/09/11 C06461858
III. IONIC EQUILIBRIUM OF TIE ATMOSPHERE
31
A. Eday Diffusion is Disreuarded 31
1. Equilibrium Conditions 31
2. Required Equilibrium Time 32
3. Ionic Densities 33
4. Account is Taken of the Inegglities of Positive
and Negative Mobilities 36
5. The Case of Radio-Active ims 39
a) Computation of Concentrations 39
1.1
b) Neutral Radio-Active Nuclei
c) Equilibrium Between Radioactive Small
and Large Ions 43
Granulometric Distribution of the Activit::
of Natur:).1 Aerosols 411
B. Introduction of Eddy Diffusion Coefficient, KAWANO's
115
Theory
1. Ionic Density 45
2. Other Electrical Parameters 1t5
IV. CONCLUSIONS
50
References 51
Figures nos. 1 to 9 after 53.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
ACTION OF RADIOACTIVITY AND OF POLLUTION UPON PARAMETERS
OF ATMOSPHERIC ELECTRICITY
by J. BRICARD
ABSTRACT
Placing ourselves at an altitude of several meters above the
ground, in order to avoid the perturbations, we recall the charac-
teristic...elements of radioactivity and of atmospheric pollution.
It is shown that their actions are in the first case the formation
of small ions, in the second case their disappearance and the
formation of large ions. From that we deduce the ionic density of
the air under given meteorological conditions, studying separately
the ions, corresponding to the recoil atoms, which come from radio-
active disintegrations in the troposphere and in the lower strato-
sphere.
Finally we introduce the relations between the radioactivity
and the other parameters of the atmospheric electricity, close to
the ground and in the free atmosphere.
I. RADIOACTIVITY AND THE INTENSITY OF IONIZATION
I. Intensity of Ionization.
We call Intensity of Ionization (or q) the number of small
ions of each sign (air molecules, having lost or attached an electron),
created in one cubic centimeter of air per second. It is hence a
fundamental parameter of the atmospheric electricity. Disregarding
the action of cosmic rays, which produce continuously about two pairs
of ions per cml of air per second at the sea-level, we can practically
say that at this altitude the natural radioactivity of the air is
responsible for 80% of the intensity of ionization.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
I' 2.
We call one Curie the quantity of a radio-element, producing
3.7 x 1010 of disintegrations per second. If we know the concentra-
tion of a given element in the air, expressed in Curies, for instance,
It is easy to deduce from it the corresponding intensity of ionization
in the case of a disintegration producing Alpha-Rays. The calcula-
tion is much more complicated and not always possible in the case
of disintegration producing Beta- and Gnmma-Rays (Section I.5b).
One Roentgen is the quantity of radiation, which per cm3 of air at
00 under the pressure of 1 atmosphere generates a quantity of elec-
tricity of each sign equal to 1 esu or 2.08 X 109 pairs of ions.
In spite of its importance it does not seem that the
measuring of q by the method of ionization chambers were satis-
factory as a whole. (Difficulties in measuring the ionization of
particles a because of the recombination in columns and the absorp-
tion of radiations by the wall-effect. Necessity to introduce in
the ionization chamber not only air, but also the aerosoiXs it
contains, responsible for one part of the radiaactivity a, and
contributing by their charge to the saturation current. Absorp-
tion of p and Ton the walls of the chamber, etc.)
In spite of the.improvements proposed (very thin walls of a
known absorption
, *, double-cage chambers L. 2:7))we know
but a few direct neasurements of the intensity of total ionization.
The instrument we applied for our calculations (double-cage) has
not given so far sufficiently reliable results to allow us to use
them at the present moment.
Thus, we are generally limited to indirect estimates of q,
at least as far as a are concerned, made on the basis of the
*) Numbers in brackets refer to the list of references.
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 006461858
3
content of radioactive products in the air and in the ground.
2. Intensity of Ionization Several Meters Above the Ground.
We shall suppose that the total intensity of ionization is on
the order of 10 pairs of ions per cm3 ' and second, 20% of which
are of cosmic origin. It is then the production of 8 pairs of ions
per cm3 per second that we attribute to the radioactivity of the
air and of the ground.
The radioactivity of the ground is about 3.5 pI/(cm3sec);
consequently the fraction of intensity of ionization, due to the
radioactivity of the air, amounts to 4.6 pI/(cm3sec). These
values are divided in the following groups, according to their origin
(Hess 1732:
Table I
-Radioactive substances of the air:
a 4.4 pI/(c1n35ec)
P 0.03
Y 0.15
4.5A pI/(cm3sec) (min 1.4
(max. 13.5
-Radioactive substances of the surface soil:
0.3
3.2
3.5 pI/(cm3sec) (mm n 2
(max. 6
We see that in the case of radioactive substances of the air
the radiation a plays the biggest role) while Yis the most impor-
tant in the case of radiation from the ground.
pproved for Release: 2017/09/11 006461858
Approved for Release: 2017/09/11 006461858
4.
Above the oceans the total natural radioactivity is reduced
to a few hundredths of its value above the ground.
3. Intensity of Ionization in the Altitude.
a. Radiation from the ground.
We see (table I) that the natural radioactive radiation
of the surface soil, almost exclusively of origin, plays an
important role in the lower layers. The following table (Israel.)
indicates its variation as a function of growing altitude.
Table II
Altitude above the ground lm
10m
100m
500m
1,000m
% of radiation at the ground 97%
83%
33%
2%
0.1%
We may suppose an exponential law of absorption of this radia-
tion in function of the thickness of the air-layer traversed. Let
141,be the coefficient of absorption, supposed identical for all
radiations. The intensity of idization is proportional to the inten-
sity of radiation in one given point. If we call cliol its value in
the immediate vicinity of the ground, we shall have in the altitude
Z:
qlz qlo exp(-If z). (1)
The coefficient/. is in the order of 8 X 10-3 m-1.
b. Derivates (Daughter-Products) of Radon and Thoron in
suspension in the air.
In the case of natural radioactivity the distribution of
concentration of Th and Rn (we neglect the presence of Actinon),
pproved for Release: 2017/09/11 006461858
Approved for Release: 2017/09/11 C06461858
5
as well as that of their derivates, is connected with the state of
turbulency of the air. Taking a soil with average internal charac-
, (tclily)
teristics and a coefficient of turbulentfdiffusion K, independent
of the altitude, of 8 x 104 cm2 sec-1, we find through calculations
Z-41 for Rn and Tn atria concentrations at the ground level of
158 X 10-18 and 174 X 10-18 c/cm3, respectively.
Difficulties arise, if we want to calculate the concentrations
in various altitudes, due to the disintegration of the various
daughter-products of Tn and Rn (it is necessary to know the state
of equilibrium mother-daughter products), and due .to the. attachment
of the daughter-products on the aerosols in the air, due to their
coagulation and to their disappearance with time.
On the other hand we have to make a distinction between Rn and
Tn. The first-one, whose half life-period is long (4 days), dis-
integrates slowly as it raises higher, while Tn (half life-period
10 sec.) and the ThA (period 0.2 Sec) disappear in the vicinity of
the ground. Thus, in higher levels only ThB remains (half life-period
10 h). In the altitude Z above the ground the concentrations of Rn
and ThB in the atmosphere are given respectively by the relations:
.1 1
Co exp (-z A 2 K-2)
(2)
where Co represents the concentration of each on the ground level,
K the coefficient of turbulent diffusion, and A the radioactive
decay constant, either of Radon or of ThB.
To simplify the reasoning, let us suppose that there exists a
radioactive balance at any altitude between Radon on one hand and
ThB on the other, and their daughter-products. This, of course, is
very approximative, for if there actually exists a radioactive balance
between Rn and RaA (3-t- minutes period), it is not so for the other
approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
6.
daughter-products, at least next to the ground. This is shown in
the studies on the decrease in function of time of disintegration
products of Radon, captured in the form of ions or aerosols. This
is generally the case for ThB and ThC.- With the simplification we
see that every disintegrationcX of Rn carries along simultaneously
2 C,. (RaA, RaC), and 2j< and 23, (RaB, RaC), and that every disinte- '
gration (5 and ' of ThB brings along simultaneously lr and 11
(ThC) and 1CN (ThC).
In order to obtain the corresponding value of intensity of ion-
ization, it is necessary to know the number of pairs of ions, produced
by each kind of disintegration, and to calculate the total at every
altitude. This number is well-known for the c) , which is monocinetic.
It is poorly determined, however, for f,f7: and V , whose spectra of
distribution of energy we know little about at the present time.
Table III represents the results, indicated by Israel, for a
coefficient of turbulent diffusion K 2 8 X 104cm2/sec and supposing
5 a middle-value of 2 X 10 pairs of ions through disintegration "X,
and 2 X 104 pairs of ions through disintegration 3 and These
values are supposed the same, independent of the source of radiation.
Table III
Altitude km 0 0.1 0.5 1 2 3 4 5 6 8 10
q from radio- 7.6 5.1 3.8 2.7 1.5 0.9 0.5 0.3 -
activity
It will be noticed that the values, indicated for the vicinity
of the ground do not concord perfectly with those of Table I. This
Is explained by the very approximative mode of calculation used.
c. Lower Stratosphere
Fig. 1 represents variations of intensity of total ionization
pro ed for Release. 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
7
in function of the altitude. We see that it begins by decreasing,
passes through a minimum at about 3 km, increases again, and above
several kilometers the effect of the radioactivity becomes negligible
as compared to that of cosmic rays. The effect of cosmic rays, very
weak in the vicinity of the ground, increases progressively with
increasing altitude up to about 12 kilometers, passes a maximum and
decreases then for the higher altitudes.
In the lower stratosphere, between 10 and 20 km of altitude,
I
we find RaD (period 22 years, source of p and 1. ) in very low
quantities on the order of 10-19 c/cm3. We find, in addition, radio-
active elements originating in the action of cosmic rays upon the
molecules of the air (principally Ar). Among them are Be., and P ,
32
which will be used later. The first one can also originate from
atomic explosions. We find about 5 x 10-19 local c/cm3 of Be7, and
5 X 10-21 local c/cm3 of P' 32 . As in the case of RaD, the resulting
intensities of ionization are negligible as compared to the effects
of cosmic rays.
4. Artificial Radioactivity.
With the exception of quite extraordinary conditions (vicinity
of an nuclear station, or in the period afternuclear explosions L7551)
the average content of artificial products in the air is now 2 X 10-18
c/cm3 of sources exclusively of 13 and. .e1 . This corresponds to inten-
sities of ionization in the order of 3 x 10-3 pI/(cm3 sec). In other
words it is negligible as compared to natural radiation, except per-
haps above the ocean, where the latter one is reduced to a few hun-
dredths of its value above the ground. The situation is the same
in the stratosphere layers, in spite of the accumulation of disinte-
gration products manifesting itself there. At about 20 km of altitude
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
3.
-17 -13 , "'
we find maximnM concentrations L752 of 10 and 10 cm ' -3, the
tl ' c;
i 3
i average concentrations being 104 and 10 times weaker. Thus the
i
effect is negligible as compared to that of the cosmic rays.
The situation is not the same, if there is an accumulation
of these products on the surface of the ground after precipitations,
sedimentation of dust etc. According to Israel L-- 41, a rainfall
of 10 mm containing 10-13 c/cm3, if all the water remains on the
ground-surface, would give in its vicinity an intensity of ioniz-
ation on the order of 75 pI/(em3 sec). However, this is considerably
minimized by the disappearance of water in the ground, .and the
effect is partially masked by the decrease of Radon exhalation durinT
rainfall. Thus, it generally cannot be observed, yet it could
become observable on a waternroof ground.
5. Qualities of Small Ions.
General Remarks. The small ions are generally present in the
air in the order of a few hundreds per cm-, the concentration (n')
of the positive ions being by some 20,6 higher than that (n") of
the negative ions. Their difference (n' - n) amounts to several
elementary charges per cm3. The concentration of the small ions,
which is normally that of 300 to 500 per em3 next to ground, can
be reduced to less than i0'/. of its value in highly polluted atmospheres
and in the clouds. It increases with increasing altitude..
Physical characteristics of the small ions, whose dimension
(some 10-8 cm) makes direct observation impossible, are the
following:
Mobility. This is the speed the ion has in the atmosphere
under given conditions of temperature and pressure, if it is exposed
to an electric field of 1 V/cm. Under normal conditions the mobility
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
9.
of the positive ions is in the average in the order of K' = 1.4 cm2/
(V sec), that of the nagative ions a little higher, K" = 1.9
UM /
(V sec). (These are in fact the most probable values, the real
values being dispersed around these average values). It varies with
pressure and toicoerature according to the relation:
P0
K(p,T) = ko (P0%) i
To
Electric conductivity of the air, more easily accessible for
direct measurement than the ionic concentration, is proportional.
to it and to the mebility of ions. It; is by Lhe
relation, where e stands for the elementary charge:
A= +1.:.nn" e
(2a.)
(3)
-16
On the [Tound level it is in the order CI: 0.5 X 10 - to 2 X
10-16 s-2. -1
cN- . Tt au:;ilents ienerally with increasiw; altitude
under conditions deynding essentially on meteorological. circumstances.
The figure 2, borrowed from M'ahleisen, represent several examples
j.//
of variation (Explorer II; Sagalyr and 2auchei"). According to curve
I we can consider as notably constant between the ground and 2.500
m (strong turbulence wider a marked inversion). Thio is not co in
the other cases.
The density of the vertical conduction current is given by:
i =A E = EK'n'e -4- EK"n"e = i" (4)
where E is the electric field strength.
In good weather the vertical current is directed downwards, and
we can suppose that its value is notably constant in the troposphere
and in the lower stratosphere, whatever the spot the measuring has
taken place may be. Its size is on the order of 2 X 10-16 A cm-2.
The constance of this current permits to bring into connection the
electric field and the conductivity of the air, which are two quantities
pproved for Release: 2017/09/11 C06461858
EJTT.,;q7.1'
c.
� To
�!
Approved for Release: 2017/09/11 C06461858
10.
of different origin.
, Diffusion Coefficient. Let there be dal the gradient of
dz
concentration of the small ions, positive, for example', following
4 direction oz, The number of small ions traversing per sec.
1 cm2 normal cross section at oz is equal to D' Wal D' stands
dz
for the coefficient of diffusion of the small positive ions. In
principle, the same is valid for the negative ions. The coefficient!
of diffusion and the mobility are connected by the Einstein relation:.
� e
D kT
(5a)
It being the constant of Boltzman and T the absolute .temperature.
On the average, D' and DI' are under normal conditions in the
order of 3.7 X 10-2 cm sec-Land 5.1 X 10-2 cm sec-1, respectively L-8_7.
Mean Free Path.- In molecular dimensions, even if the charges
are elementary the electrical field produced by a small ion,
animated by the Brownian movement, is sufficiently strong to polarize
the neighboring molecules. Its trajectory, which is deviated by
these charges between two succehve collisions, is not ,straight, so
that the results of the cinetic theory cannot he applied to it
any more..
If we neglect this effect, we can calculate the mobility and
the diffusion coefficient in function of the masses m of the ion
and M of the gas molecules, from the average speed of thermic
agitation of the ion and of the gas molecules and from a fictitious
mean free path A� Thus we obtain, for instance, the relation
(1st formula of Langevin):
K uae
8
(5)
41.11111111111.111111111.1.11=11.11MIApproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
11.
where k stands for the constant of Boltzmann.
If we take M � m and K = 1.5 cm2/(V sec) (normal conditions of
temperature and pression), we find A = 1.3 X 10-6 cm, a value
notably different from the value, which corresponds to air molecules
(6.4 x 10-6 cm), under the same conditions L79_7%
Recombination Coefficient. The small positive and negative ions � ,
recombine after their formation. A number of small ions of both signs
disappear thus per sec and per cm3s
dn' dn" C( n' nu; (6)
dt dt
.114
e4 standing for the coefficient of recombination, is given by the
relation:
47-2.- (12 e2 ) E
3 �ieT
1
Af5 4kT 87e2
(7)
(8)
4 E-2
6.Fe2
9
*) - exp
� 1 t 3kTcr/e:A
3
((e2
6kT
2
)
( .
- I: 6ek2T
2 2) 3/2
A
�
�������
�����
e2
6kT
where k is the constant of Boltzmann and 2 the mean free path, defined
::��4
*!'�44,14#1�
in (5), and q"-is the average speed of thermic agitation. The express-
ion (9), valid in a range of pressures between 102 and 10 5 sib is
different from that of Thomson Z-81 and allows us to calculate the
coefficient o under the various conditions of temperature�and pressure.
Under normal conditions we find (-15.7 Q(= 1.6 X 10-6 cm3- sec-1.
(9)
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Every disintegration of Rn, of Tn, and of their daughter-pro-
ducts in the atmosphere is accompanied by the appearance of a recoil
atom. Let C be the concentration of the element considered, expressed
in Curies per cm3. For Radon, for example, there will appear per
cm3 and per sec. next to the ground a number of RaA atoms in the
the same as for the other constituants. A certain number of these.
atoms is neutral (20% in the case of RaA), the others with a positive
unitary charge, constitute the small radioactive ions.
Studying the decrease of these small radioactive ions, as a
function of the time, properly sampled, we find that they consist
almost exclusively of Rak, with a very weak fraction of RaB. A more
profound study allows us to evaluate the life-span of these atoms
in the atmosphere before they disappear in a process we shall deal
with further on. This life-span is found to be on the order of 20
It is easy to determine their concentration in the air by
direct capturing, and by measuring the activity of the captured
sample. We find in the vicinity of the ground that this concentration
is on the average of 10-4 to 2 X 10-4 atoms
present time), we shall suppose that their mobility and their diffusion
coefficient is the same as that of the ordinary small ions.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
13,
A study of the products of artificial disintegration, found in
the lower stratosphere (accumulation zone) would lead us beyond the
framework of this article. Without getting involved in the details,
we can say that one part of the products of the explosions exists
Initially in the atomic form, neutral or electrically charged, as
well as all the products of fissions of cosmic origin.
Let us take a simple example, that of Be, and P32/ generated
by the fission of oxygen and nitrogen in the first case, of Argon
in the second case. The period of the first (Be7)' which changes
into Li7 stable, is 53 days, that of P32, which changes into S32
stable, is 14 days. All these are products, which are set free In
the state of atoms, and they have probably an elementary positive
charge.
Let us suppose at the altitude of !0 km L714_7 the concentration
of 5 X 10-19 c/cm3 of Be7' and 5 X 16-21 c/cm3 of P32. They remain .
in the stratosphere for a sufficiently long time so that we can
suppose that the radioactive equilibrium is reached. According to
the relation (10), in one cm3 of air and per sec. there will appear
-10
2X10-8atomsofBand as many of Li7, as well as 2 X 10
e7 atoms
of P32,- with as many atoms of S32. These atoms exist for a certain
time in a free state in the atmosphere before they are attached to
aerosols, present at this altitude (section II, 1), hrid it should
be possible to detect them (section III 5 a ).
Approved for Release: 2017/09/11 C06461858
L.ac1. la
,
Approved for Release: 2017/09/11 C06461858
14.
II. AIR POLLUTION AND LARGE IONS
1. General Remarks
Natural aerosols, the aggregate of which constitutes the
normal atmospheric pollution, are liquid or solid particles, lilcely
soluble in water, neutral of electrified, whose constitution is, as
yet, not well known and to which we will assign a spherical shape.
-4
Their dimensions are included between 7 X 10-7 and about 10 cm radius
(see Table IV). They are also called condensation nuclei.(see sec-
tion II, 3).
Condensation nuclei, appearing at the ground, are carried
((tidy)
upwards by turbulent'diffusion. They coagulate but slightly and
fall back onto the ground by sedimentation, preferably at night,
when turbulent conversion is less intensive, or they are collected
during the fall of precipitations.
They may come from the ocean (sprays), or from human
activities (combustions), and, owing to their slow falling speed
with regard to air, they are apt to be carried far from their place
of crigin. Their chemical composition may vary greatly L7161,. C17_7
(chlorides, sulphates, nitrates, ammonium salts, sodium, magnesium).
They are usually mixtures of various �matters, since they coagulate,
coming from simple nuclei initially formed.
Their concentration ranges about 10-) to 105 per cm3 at a
few meters above the ground. It decreases with the increase of the
altitude, said decrease being more or less steady (see Large ions,
electric charge, section II, 4). Figure 3, borrowed from Isragl
L741, shows, according to Wigand, Z.-18/ the relative variations of
atmospheric concentrations in aerosols as the altitude increases
pproved for Release: 2017/09/11 C06461858
" .
� : r.2?-�
Approved for Release: 2017/09/11 C06461858
15.
(measures made.by a balldon). It may be seen that said distribution
follows an experimental law, with a discontinuity corresponding to
1
a temperature inversion. Towards 8,000 m it is reduced to of
104
its value at the ground level, and it goes on decreasing as one goes
up. According to Junge 119.1, there would be an increase of the
concentrations, between 10 and 20 kin; the concentration of particles
averaging about 0.15,,u radius, going from 0.01 to 0.1 cm-3, before
decreasing again. But this concerns rather large particles and it
does not seem that, at these altitudes, smaller particles have been
numbered.
Let us finally mention the case of natural clouds, made
up of droplets of water.of someiti in diameter, a few hundreds of
which may exists in a single cm3. Some authors have mentioned, in
addition to these droplets, the presence of particles inferior in
size and much more numerous 1.720_7. Above the altitude of 6,000 in,
clouds are exclusively made of ice particles flat or elongated, over
0.5 mm in diameter. These crystals play an important part in charge
generation in stormy clouds, but this will not be discussed here.
A direct observation of natural aerosols offers difficulties
in particular with respect to their sampling. This is made either
by collision (Konimeters), a method which does not seem to be applic-
able to small size particles under 10-5 cm or about; either by
means of very fine threads (spider threads), which is possible
only for liquid particles; either by electric precipitation of part-
icles charged by corona effects, (particles with dimensions between
approximately 0.214! and 0.8iAL escape more or less to that kind of .
precipitation); either by thermal precipitation (which is only good
for solid particles); or by means of filters (one may moreover wonder
how it comes that particles settle on the front surface of the filter
pproved for Release: 2017/09/11 C06461858
.c2.�
or:
Approved for Release: 2017/09/11 C06461858
16.
and that, as a rule, only a few of them get inside the pores). On
the other hand, particularly in the field of dimensions unattainable
to an optical microscope, it will not be possible to use an electronic
microscope for liquid particles, as long as there are no means avail-
able to realize supports liable to fix impressions of the droplets
of these dimensions. In order to 'study condensation nuclei granul-
ometry, we are compelled to use indirect methods based on their
physical properties.
Their quantitative analysis in bulk in the air may be
achieved directly by sampling on filters and by chemical analysis,
or by radioactive computation. This prodeeding, justly crticized
C43.7 on account of the inefficiency of filter samplinr in some
dimensional fields, particular]y between 1.5 X 10-6 and 10-5 cm,
seems now perfected. It has been controlled thanks to the use or
extra-thin calibrated aerosols and of large natural ions, the size
of which were known. Filters with an efficiency of more than 98ro
may thus be obtained, whatever may be the dimensional field of
aerosols to be filtered.
� 2. Optical properties
Light is diffused by these particles and, at least in
the case where their constitution is known, (index, reflection,
and absorption factors), it is possible to compute a diffusion
indicator for particles the dimensions of which are given. Recip-
rocally, the measurements of the flux diffused by a group of particles,
allows to figure, at least approximately, the atmospheric concen-
tration in aerosols; and the measure of the flux diffused by an
isolated particle, correctly lighted, allows to know its dimensions
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
17.
without changing its constitution.
This method has been originally used for qualitative mea-
sures or for pollution detection, has been recently developed so as
to become quantitative L7201, L-211. It permits the access to.
dimensions bordering on the limit of the separative power of the
optical microscope and we are now extending it to a�field of lower
dimensions.
3. Condensation
These particles, in a supersaturated atmosphere, act as '
condensation germs and give liquid droplets directly observable
optically. The use of said property permits to determine their
concentration in the air.
A droplet thus formed has a radius depending on its
chemical nature, on the salt concentration or on the mixtures of
salts of which it is formed and on atmospheric supersaturation.
By measuring these droplets, under well defined conditions, it is
possible to determine the primary dimensions of the corresponding
germs in the air And to form an idea on the air pollution on the
spot where they have been taken.
The following granulometric distribution may C22J be
inferred from this:
dN
d log R
R3
(11a)
in which dN represents the number of nuclei in a logarithmic
scale of radius R, and C a constant. This expression seems
valid for particles whose dimensions are at least equal to 10-5cm
(see section II, 6).
pproved for Release: 2017/09/11 C06461858
I
tiette.A.1.7�.:=M10:3Ctilt:t!%:=4;17Nertir,4tW4
Approved for Release: 2017/09/11 C06461858
18.
4. Electric Charges
These particles can carry charges usually very low,
(ranging about one or several elementary charges), either by .
attachment of small ions, or by means of other chemical or thermal
processes, in order to give large ions. The statistical study shows
that, on an average, the atmospheric concentrations of large ions
of both signs are very close, so that the average space charge
corresponding to them must be low. However, in the course of the
various operations resulting from human activity (combustion,
condensation), very important differences between these concen-
trations, as well as a very pronounced space charge, either posi-
tive or negative according to circumstances, may appear.
Under undisturbed circumstances, the proportion
of the total number of nuclei, to that of charged nuclei, is very
variable, depending on authors L7232. it Is included between
1.61 and 5.4. It increases with the number of nuclei; the lowest
value nearest to that stated by Mme,Thellier C24_7, seems to
correspond to undisturbed average statistical conditions.
The large ions concentration, proportional to that of
� condensation nuclei, decreases also as the altitude increases, under
conditions depending on the meteorological situation; and they
disappear above 3000 or 4000 m. Fig. 4, borrowed from Sagalyn and
Faucher C25_7, represents examples of their distribution in altitude.
5. Mobility and Diffusion Coefficient.
Due to their rather great dimensions, the above mentioned
particles have mobilities and diffusion coefficients defined in the
same manner as in the case of small ions, but much smaller. The
�
7.!!:�,�
.1 I
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
19.
-2
maximum values, under normal conditions are, ranging in about 10
cm2/ (V sec) and 3 x 10-4 cm /see, respectively. These quantities,
as in the case of small ions, are interrelated by relation (5a).
It is known that, in a viscous fluid, the strength needed
to give a constant speed B to a particle with a radius smaller than
10-5 cm is expressed as follows (Stokes-Cunningham. law)
6 1r R7
F (12)
1 -Osb/pR
2 is the air viscosity (at normal temperature and pressure
1.7 X 10-4 cgs), p is the atmospheric pressure expressed in cm of
=
mecury)and b a constant, (b = 0.000617); R is the particle radius.
According to the mobility definition, we may write:
K
1
67R7
ne ( 1-f- );
pR
(13)
e meaning the elementary charge and n the number of chariT,es 'carried
by the ion, said number being low and usually equal to the unit
(see further down).
The following table IV gives an idea of the large ions
mobilities and dimensions, derived from formula (8), a Unit-charge
carried by the ion being supposed.
Table IV
Small ions 1,0 > k > 0,01 cm2 V-1 sec'
Average large ions 0,01 > k 0,001
Langevin ions 0,001 ) k > 0,00025
Ultra large ions k 0,00025
6,6 10-5 cm and that no com-
putation is also satisfactory for the lower values of radius.
b. Introduction of the Mean Free Path of Small Ions.
Suppose that Xis the mean free path of small ions, stated
by relation (5). Suppose that is the average distance from the
particle surface, with which the small ion had its lust collision.
It is supposed that everything happens as if said distance would be
constant, i.e. as if the collision had taken place at the distance
R-4-A from the large ion center, and that the thickness shell,A,
may be considered as a void space. A first approximation, (Arendt
and Kallmann (45_7) used by Lassen (46_7, consists in taking A-
. In a more accurate way, (Smolukowsky L-35j), we shall take
_
3 R A
(27)
which gives L= ) for very srma values of R, and Ls. = .A for great-
2
values.
It may be considered that, in the thickness,zoneA ,
between both surfaces, particles (in this instance small ions), move
as in a vacuum and are impelled by thermic agitation speed v.
Suppose that n A is the small ion concentration on the outside
surface of the shell. The ion flux reaching the particle (large ion
or nucleus) is:
Amommm�MIMIIIIIIIIMMINIMMEMAoproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
In saying that the above relation is satisfied and that, very
from the nucleus, ionic densities have the same n value, we get
If we disregard the electric image, we get a good approximation
pproved for Release: 2017/09/11 C06461858
, � Approved for Release: 2017/09/11 C06461858
in which k represents always the Boltzmann's constant.
We find especially for/:
A
2
R v
R2
1+�.--
D R
(34)
30.
We see on figure 6, that the curve marked II, new theory
calculated from (33) and (34), confirms the experimental points of
Keefe, Nolan, and Rich, in a satisfactory way. Another argument in
their behalf, is that relation (8), concerning the recombination
coefficient of small ions, and equally proved right by experience,
has been obtained through an argument identical to the above argu-
ment.
If r is greater than some 10-6 cm, relations (33) are
simplified and become identical to those of Lassen L7461. For
example we obtain for
o
o ;
vR
(35)
-5
and if R is greater than 10 cm, we obtain relation (24), save for
a few hundredths.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
---Krtn,;,:tpme4
�.r....-
;�� � �
�:, :
� � ��
31.
III. IONIC EQUILIBRIUM OF TiE ATMOSPHERE.
We have studied, in chapter I, the radio-active origin of
small ions and, in chapter II, the pollution characteristics, keeping
strictly within the purpose of the Atmospheric Electricity viewpoint.
We intend to study here the combined action of radio-activity and
pollution, that is small ions versus condensation nuclei.
The consequences of the relations obtained, which are valid
through the thickness of the entire atmosphere, will be studied circum-
stantially in the lower atmosphere, where experimental results are
comparitively numerous. We shall state, furthermore, what may be
their possible application to the problems of the higher atmosphere.
A. Eddy Diffusion is Disregarded.
,
1. Equilibrium.ConditionS. We suppose that the small ions, k4:�'1i:s4,4t
:"!�� e,
created in the air, disappear solely by fixation on neutral nuclei
or on large ions having an opposfte sign or the same sign; placing
ourselves in a quiet atmosphere, free from turbulence, we disregard
t7,
the nuclei coagulation (section 11.6). At a given moment, at first
kr1:4,r
r,
we suppose that charges are symmetrical and that all the nuclei have
the same radius (monodispersed medium), calling "Np" the number of
nuclei with the charge " i-pe" (large positive ions); and"N:the
number Of electrically neutral nuclei.. On account of the symmetry
of the problem, the number per cm3 of nuclei bearing the charge
"- pe" is also N. Under these conditions we may write:
dn = q - CI( n2
tx-
�rp ( N
/Lap ts-- 2p
p 1 I
d(nE) �
- K (36)
dz
� pproved for Release: 2017/09/11 C06461858
- � ,
Approved for Release: 2017/09/11 C06461858
32. '
19)//)2p are defined by relations (33...). In the vicinity of
the ground, the term depending on the re-combination coefficientOk
will be disregarded. This would not be allowable in altitudes where
/11
quantities 0No
1N1 4-/IN are much lower. We shall
22
also disregard the term Kn dE .
dz
We write for large ions:
dN
= n//,
- N
/ 1,p-1 p-1 np p
dt
�
(37)
We shall state that an equilibrium exists between the large
ion production and their attachment on the nuclei, consequently:
dn :0 dN =0
dt dt
( 38 )
2. Required Equilibrium Time.
Let us suppose, in order to simplify, that there are no
large ions having a charge greater than 1 elementary charge. If we
consider the concentration of nuclei and large ions as constant, the
concentration of the small ions at the time t will be written:
9
n (t) = N exp (- )t) (q//Y) L71 - exp (-/t)1 (39)
If the total nuclei concentration (i.e. Z) is constant, the
concentration N of the large ions of both signs will be:
N(t)
- exp (-pt)_.7 N exp (-pt)
1-2 o �
- exp (-pt)1 No exp (-pt) (40
o is the value of N at the time t � 0 and N the equilibrium
pproved for Release: 2017/09/11 C06461858
concentration.
8= .
Approved for Release: 2017/09/11 C06461858
1
0N0 +7,21
33.
represents the average life time of a small ion, In a free state,
that is the average time elapsing between the moment when it appears
and the moment when it settles on a condensation nucleus.
In normal atmospheres, that quantity is included between
20 and 50 sec.
The quantity:
1 1
T =
(711- 2 /2 )n
21 / 0
(1i2)
represents the equilibrium time constant. Under normal conditions,
it averages a few hundred seconds L74:7. A direct measure of the
average age of large radio-active ions, in the vicinity of the grounds,
gives a value averaging 15 minutes. This represents the time of
contact between large ions and small ions. The result is that in
this time, as a rule, the ionic equilibrium corresponding to relations
(38) is nearly reached) a few hundredths still missing.
3. Ionic Densities.
wherefrom:
According to (36) we shall write to the equilibrium COI
q = nigoN0
N
2p p
(43)
It is easily established, according to (37) and (38), that;
/)
o ,/"" 111 r-'1,1 p-1
= N
P 0
.4 21 /122 752p
Z No (1 -+ 2
:EI aP)*
1,41-
No a
(W)
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
34.
These are relations whose numetic computation is in progress
and which Permit the determination of the distribution of charges
in terms of the nucleus.
We may state that if R does not exceed 2x10-6 cm, particles
carrying a double charge do not exceed 1% of those carrying only
one charge; that particles carrying three charges do not exceed by
2% in nuMber those carrying a double charge.
We find the re)letition of charges attached to particles
with greater than 10-5 cm in (37) and (38). In particular, it is
shown, that the maximum charge attached to cloud droplets is in the
order of a few tens of elementary charges.
Relation (.44) may be written:
11 (1-- 1p
/21 2p
or, accordinG to (25)
exp �
N m N exp
PI?
exp (m/2) - exp (-pt7/2)
:N exp
P ?
If R exceeds 10-5 cm, the quantity
(-P29/2)�
P1)
exp - exp
P?
is very near the unit, so that we may write (Fuchs Z-33_7) for every
sign:
22
P e
N N exp ( ).
o 2 RkT
The distribution of charges therefore follows, when the
particles radius is not smaller than 10-5 cm, or so, a Boltzmann'S '
pproved for Release. 2017/09/11
Approved for Release: 2017/09/11 C06461858
35.
law, as regards electrostatic energy, according to the hypothesis �
expressed by Nolan, Keefe, and Rich L744:7, (section II, 8a), which
may be considered as an approximation to the results stated above.
If we disregard particles carrying several charges (i.e.
.-11Y % ) and if we suppose that ( / = 2, (in fact the
11) : 21
average of this ratio amounts to 1.7 according to table (VI))we
find, according to (24) and (43):
q = nZ/3o =
nz
4IrDR
(R,o)
(46)
R is the average radius of nuclei. This is Schweidler's formula
which connects density of small ions and condensation nuclei with
ionization intensity.
In a more accurate way, we shall write, according to (43):
=
+/..) ap
1p 2p
1+ 2 (
1)-1
8p)
(47)
71 DR
I (R) (48)
Figure 7 represents [49_7 the variations of function I (R)
in terms of R, in the case as expressed by relation (23a) and in the
case as relations (33) are used. Ue see then that, except for very
small values of R, function I (R) is very close to the unit.
Introducing the atmospheric electric conductivity, relation
(47) is written:
AL 2 q I(R) Ke 2q Ke
4/ 11
4i DRZ
(49)
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
36.
Figure 8, borrowed from Sagalyn and Faucher L-25_71 rep-
resents ionic density variations, computed according to the above
relations. The points marked show the results of measures made in
the exchange zone, which had a thickness of a few km above the c;roUnd,
Let us now consider the case of a medium whose granulo-
metry follows relation (17). Relation (47) is written as follows:
q n
re"
.47
3 f(R) dR = Cu Z
(50) �
C being a characteristic constant of the (;ranulometry for
a certain ioni ation intensity, that is for a certain stated atmos-
pheric radioactivity; we shall then write:
-1
n x Z te
= C ,
an expression which is experimentally proven L727_7, L728:7.
4. Account is Taken of the Inet;alities of 1-'o4itive and Nef:ative
Mobilities L-34_7.
Relations (47) will be expressed thus:
41TD'RZn' 4 IT IP iizn"
q
with
I (R, p')
(R, p")
n"
4 it-D,Rz 11.`ir D"RZ
n' D" K'
it results that:
n" D' K"
(52)
(53)
(54)
According to the above relations, the space charge constituted
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
by small ions, will be written as follows:
n")
- D"
4 1:?Z D' D"
37.
(55)
in which R is always the average radius or the large ions.
This quantity represents a positive space charge;
for an average ionic density near the ground it averages to 50 to
100 elementary charges per cm-), in accordance with the results which
may be computed from direct measures of ionic concentrations of
both signs.
As a rule, this represents only one part of the space
ehRrge, which, while preserving usually a near-by order of size, may
be inversed in sign. Norinder's L750.1 results, valid at 8 or 9 m
above the ground, show a negative charge (400 elementary charges per
=3).
Scrase C512. finds in case of marked turbulence, a charge
always positive (under these conditions the electrode effect is
masked); and negative in the first 5 meters above the ground, if the
air is quiet; the average value measured ranging about 200 elementary
charges per cm3. From here it may be inferred that there is a surplus
of charge born by other particles and especially by large ions in
excess over that corresponding to small ions; and this may reverse
the sign of the whole.
On the other hand, positive and negative conductivities
are such that:
if
q
K' n'eZ'
411 D' HZ
q kr"
= K" n"eZ
4-11-D"Rz
(56)
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
38.
It results from relation (54) that:
1\+' )" 9
(57)
According to relations (4) and (56) we find that the
corresponding electric field of the earth is really proportional to
tho pYl_lati.Gn and invercly -;:rororchal to
the ionization intensity, which corresponds to the observations
and permits particularly to explain the field variations related
to radio-activity, on the ground as well as in altitude [55].
Finally, according to (23), (37), and (38):
N' N" I (
_
N" I L , -(p A-1)]
p-F1
p4.1
Particularly:
(58)
(59)
Relations (54), (57), (59) confirm usually stated experi-
mental results. The following table gives a comparison with Mme.
Thellier's experiments:
Theoretical Values EAperimental Values
n' k" n'
- 1.25 = 1.24
n" k n"
/1\ = 2-
= 1.42 X 10-4; Au It 1.40 x 10-
N'1 N'1
- 1 = 1.03
N" N"
1 1
r:
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
39.
According especially to relation (59), the space charge
born by large ions should be zero, which results seems inconsistent
with the results recalled above. However, we must not forget (section
III A, 2) that the ionic '-is usually reached only to a few
hundredths. Let us take 5%. It corresponds to a possible excess of
large ions of the same order. For a slight pollution (2000 large
ions/cm3), 100 elementary charges born by large ions may result) .
this being all the more marked when pollution is higher. We must
furthermore add that artificial Pollution causes important charges
of a preponderant sign Z. 6_/, which have no time to be neutralized
by natural ionization, and whose action, superposing on that of
natural pollution, alter ionic\ /equations.
5. The Case of Radio-Active Ions.
a. Computation of Concentrations. The following reasoning is
prevailing and valid whatever the nature of the radioactive body present
in the atmosphere may be, providing that elements of the range of
molecular dimensions be present and not aerosols of larger dimensions.
It will be applied to the particular case of the Radon near the
ground, which seems to have been the more accurately studied until
now.
Supposing that:
q = N (6o)
A Rn Rn
3
is the number of RaA atoms appearing per cm of air and per second,
N meaning the atmospheric concentration of Radon and X ' its
Rn En
radio-active constant. 'de will write relation (36):
pproved for Release: 2017/09/11 C06461858
dn
A
dt
Approved for Release: 2017/09/11 C06461858
q -Dcrin - n
A �A A, o
40.
A/ 1 (
>2p) Np 11?ir s')".
nA is the atmospheric concentration of small RaA ions, and N4 that
of the nuclei having attached RaA atoms.
Not knowing the exact value of the diffusion coefficient
of neutral RaA atoms (mentioned in section (I,6)), which appear when
the Radon disintegrates, we suppose as a first approximation, that
every radio-active ion carries a positive elementary charge. The
problem is then the same as that of the attachment of ordinary ions
on condensation nuclei, with the difference that this concerns only
positive ions.
Due to the fact that small radio-active ions have the
same mobility as the small normal ions of the atmosphere, it results
from relation (7) that they have also the same recoMbination coefficient
as the small negative ions and we may, under normal conditions, dis-
regard the quantity am< nnA.
The have the same meaning as in the previous section and
the expression N takes into account the formation of radio-
active particles having a multiple concentration charge NpA. The
limit concentration reached at the equilibrium will be given by
dnA - 0; that is, according to notations of section IIIA, 2
dt
and disregarding multiple charges:
qA
n = � . (62)
A
A
The quantity gi 77- represents, as in the case of
i)
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
�
'ew
relation (41), the average interval of time between the appearance
of a radio-active atom coming from the Radon, disintegrated or not,
and its attachment on a particle; it ranges therefore to about
20 to 50 sec, the same as for small normal ions.
This is, in fact, the order of magnitude corresponding to
Renoux' direct measures L7121, taken close to the ground (study of
the decrease of disintegration products of small ions, directly collected
in the atmosphere during a very short period). RaA atom concentrations,
corresponding to the conditions of relations (62), are of the same
kind as those mentioned in section(I.6) for i.-values consistent
with experience.
Relation (62) may also be applied to the computation of the
concentration of stratospheric recoil atoms. At an altitude of 20 km,
we may suppose for them: D l cm2 sec-1; R � 10-5 cm; and Z = 0.1
per cm3. We find that ranges at about 10-5sec-1, the attachment
time for every present aerosol particle of these liberated atoms
ranging at about 10 sec., about 3 hours. If we take the case of
Be7 and P32 considered in section 1-6, whose period is long com-
pared with this attachment time, according to (62), corresponding
concentrations would amount to 2 X 10-3 and 2 X 10-5 of free Be
7
and P atoms per cm3 of the atmosphere, respectively.
32
b. Neutral Radio-Active Nuclei L7293 The whole of the Radon
decay products is thus either free, in the form of small ions,
mainly constituted by RaA, and by RaB and RaC in very small quantities,
or in the form of radio-active condensation nuclei, electrically
neutral, (coming from large ions originally negative),, or of. large
radio-active ions (coming from neutral condensation nuclei), or
e
� �<
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
42.
axed on dust particles.
Radio-active neutral ions can also appear by neutralization
of large radioactive ions, by small negative ions and disappear
by attachment of small ions of every sign. We will disregard, in a
first approximation, these minor reactions.
If the disappearance of large radio-active ion and neutral
nuclei, exclusively of NA and N0,4 concentration, is exclusively
attributed to their decay, we will write, for instance for the
second concentration:
dN0A
A7
n N" - ;\ N.
A 21 A 0
dt
(63)
N" is the concentration of large negative ions. The same relation
will be obtained for N.
If we suppose that there is an equilibrium:
dNA dN
0A
(64)
and supposing a probable atmospheric ion concentration, we find
for n = A 1.5 X 10-4cm2 (section 1.6) that Nak = 5 x 10-4/cm3 N=
A
4 x 10-4/cm3.
This corresponds to the experiments (section 11.7).
However, (secondary reactions mentioned above), large
negative radio-active ions should appear, originated by attachment
of small ordinary ions, on neutral radio-active nuclei. In spite
of all our efforts, we did not succeed in detecting the presence of
these large negative ions in the atmosphere, which leads us to
suppose that such a mechanism does not exist. It is, however, possible
to realize it by the corona effect. An explanation could be the in-
sufficient contact time between natural aerosol particles and Radon
pproved for Release: 2017/09/11 C06461858
�,:t4
t � II ..�. .�
Approved for Release: 2017/09/11 C06461858
43.
� products.
c. Equilibrium Between Radioactive Small and Large Ions.
(10_7,fil11-11.31supposathatz.is the concentration
of RaA atoms, corresponding to these two kinds of particle. Due
to the smaN RaA period (3 min.), it may be supposed that the
raaio-active equilibrium is obviously reached between the Radon
and the RaA resulting from its decay (10 minutes are indeed
necessary to realize said equilibrium to 10%). We shall there-
fore write:
q = ;\J012:-- (nAA-ZA ).
A '
and, starting from (62) and (65), we find that:
Z =q
A A
from which:
nA
%3
A
nA
We see that the ratio is independent of the
(65)
(66)
(67)
quantity of Radon existing in the Aair.
Relation (67) shows for the RaA (period 3 min), near the
ground, under normal conditions, ((3: 4 X 10-2sec-1), that ranges
A
at about 1N which corresponds to the measures. These relations are
also valid for stratospherical decay products. For Be7 and P� with
i4
10-5sec-1, orders of magnitudes of 0.02 and 0.06 are respectively
found for n/Z.
r:
L1',1
.t.
$44,
. 51.
Te;?;.. 1:.�!e4
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
44.
d. Granulometric Distribution of the Activity of Eatural Aerosols
C462, 01-9.1. First, let us go back to the case of the attach-
ment of small ions on natural aerosbk) but in a polydispersed
medium. According to (47) and (48), we may write:
q(R) = 41rDn Rf (R)
1(B)
(68)
dZ
f(R) ---- represents the granulometric distribution of the medium
dR
as given by the relation (17). Function q (R) is represented, in
relative values, in fig.9 for two values: Rom. 5 x 107 and Rt., = 10-6
cm of minimum radius R, corresponding to inferior limits, reasonable
for large average ions: the maximum radius was chosen equal to 1 .
The distribution, thus computed from relation (68), is practically
identical with that computed by Lassen L746:7 from relation (35)
for Rd) 10-6cm.
On the same figure, a dotted line shows the distribution
obtained by making I (R) z CC in relation (68). We may see,
especially for Ro = 10-6 cm, that the corresponding distribution is
very similar to the former, and that, due to the accuracy of
measuring methods now available, this approximation is largely sat-
isfactory, which justifies Mthleisen and Hollis L727.7, L-281
arguments L7I (R) =
We derive from the curves that the average nuclei ( R 111:
- ,
2.5 X 106 cm) attach 78 to 80% in the first case and 83% to 86% in
the second; the remaining, i.e. about 15 to 20%, being attached by
particles of larger size. These results define the role of these
particles, the nature of which has been discussed before L7311....7.
Relation (68) is general and applies to small radio-
pproved for Release. 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
45.
active ions derived from the Radon; for instance, attached to aerosol .
particles; n is to be replaced by n and q by q (R), which represents
A A
the accrued granulometric distribution of the activity attached to
natural aerosol particl._s. Points marked in fig. 9 represent the
experimental results directly obtained, corresponding to table V.
We see that categories (1) and (3), table C, are well placed, (points
marked c and d), with respect to the computed curves in fig. 9.
The experimental device used in these investigations does
not permit to know accurately the value of the radius corresponding
to class 2, which may be either a or b on fig. 9. However, even if
we take the farthest, we are again in the neighborhood of the computed
curve. We may then conclude, in consideration of the slight
quantities measured and of the difficulties encountered in such
experiments, that the experimental points thus captained, constitute
a first checking of the theoretical consideration stated above. They
are also in agreement with the theoretical distribution foreseen by
Lassen L7461, concerning the attachment of Thoron by-products on
artificial aerosol particles.
B. Introduction of Eddy Diffusion Coefficient. Kawano's Theory
152/ C53-7.
1. Ionic Density.
In order to know the vertical distribution of ionic densities,
we must introduce the eddy diffusion coefficient into the equilibrium
equations. This allows to determine the distribution of the various
electric parameters, depending on the altitude. Let n be tte
concentration of small positive ions, for instance, supposing that,
as in (36), charges are symmetrical and that K is the eddy diffusion
coefficient. The rate of variation of n with altitude Z will be
Approved for Release: 2017/09/11 C06461858
4
Felk
�:0�
H
Approved for Release: 2017/69/11 C06461858
expressed, on account of eddy diffusion, as follows:
dn dn
( K ----)
dt dz dz
46.
(69)
The contribution of ions, coming from lower parts, adds
itself to the production of ions by the various ionizing agents,
considered in chapter I. Let q(z) be the corresponding total ioniz-
ation intensity. Equations (36) and (33) become, at altitude Z*:
d.
dz
dn
(K ----) � q J3 Zn - + n2.
d(nE)
dz - �Tr (
(70)
Account may be taken of possible variations of K with the
altitude (Miline L-541), which complicates the computations. We
will merely consider K as constant (Kawano) and disregard possible
variations of the conduction current, as well as the small ions recom-
binations, which leads to relation:
d2n
K q = R Zn.
dz`
( 71 )
The ionization intensity is the sum of 3 terms:
qz qz qz, qz -
(73)
The first term)-z1-
oi , represents the action of the (-radiation
from the ground. At z altitude, it is given by relation (1). The
second term represents the action of radio-active gases and of their
active decay products suspended in the atmosphere. By limiting our-
selves, in a first approximation, to the Radon daughter-products and
*) K is not well known close to the ground. However, we shall suppose,
that the value of K as well as that of dn/dz is much smaller at ground
level than in some height: This justifies the equations (37) and (33).
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 006461858
by reverting to the reference level h, we will write, according
to (2):
q2(z) = q exp - (Z - h.)7.
2h 1 K
f
(73)
47.
where A is the Radon radio-active constant. (It would be easy
to take into account the other radio-active bodies, possibly present.
They would reveal themselves through a sum of terms of the above
type).
Relation (70) will then be written under the form mentioned
by Kawano:
dn
K q exp(7A4z) q exp -
dz2 10 211
K
(z-h)1 A- q
3
. Zn; (74)
q is the cosmic radiation contribution, supposed constant, in
3
terms of the altitude.
Z n�; (74)
The integration of the above relation has been made by
KAWANO, in the case when product Bz is constant with the altitude.
It is evident that this condition is in contradiction with the other
conditions brought in relation (74), (other characteristics variable
with the altitude), as well as with experience, (section 11,1).
Therefore, the expression of n thus obtained can only be very
approximative and would most probably be improved if Z would be
given an exponential form. However, with the following limit con-
ditions:
n 2 nh at z a h
n = n
at z =DO
pproved for Release: 2017/09/11 006461858
Approved for Release: 2017/09/11 C06461858
8.
we obtain:
=
qlh q20
_BZ-A z - K
exp (z_h)J 4
K
_BZ
exp
J-1.2
K
(z-h)
20
2exp (
q
.1Z-K BZ
(75)
This expression allows numerical calculation of the ionic density
with these simplifying hypotheses and according to (3), the conduc-
tivity, resistivity, etc., at various altitudes.
2. Other Electrical Parameters.
By calling E the earth's electric field supposed vertical
and the Charge density at altitude:
(76)
dz.
or, according to (4):
s iorA2.
dz
(77)
i is the density of the vertical conduction current.
Starting from (3), (75), (76), and (77), we may therefore
bring in the pollution and radio-activity on all other atmospheric
electricity parameters, and especially on the space charge and the
electric field. Let us write q20-1- (13 = 012 andjus 0, which means
to disregard ground radiation absorption (tab:Je III). Noting that
jk= 2.09 X 10-6sec-1 is negligible compared with 3Z (whose order
of magnitude under low pollution is nearing 10-2sec-1), we obtain,
according to (75), (76), and (77), a relation between the electric
field E and space charge density at the height h, which is written
pproved for Release: 2017/09/11 C06461858
( 3
Approved for Release: 2017/09/11 C06461858
n t-9-9
h Z
(78)
4 2
At zero altitude, if we take K = x 10 cm sec, B Z
10-`sec , qh m-3 = 10 pI c sec-1 , and n = 1000 cm-3 according to
0 -1
(46), we find that E ranges about 0.3 esu/m, i.e., about 100 V m-1,
normal 'size range. Relation (78) allows to explain, in a satisfactory
way, the electric field local anomalies.
I
pproved for Release: 2017/09/11 C06461858
,
Approved for Release: 2017/09/11 C06461858
QtJITIt:AK
IV CONCLUSIONS
50.
Although the relations established in the course of this
report are general, their application to problems of atmospheric
electricity is restricted, as regards the equilibrium between small
ions and natural aerosols, at altitudes higher than a few km. Above,
ionization intensity of radio-active origin and atmospheric pollution
play a negligible part, compared With ionization intensity of cosmic
origin and re-combination between small ions. In other words, relations
(70) and (72) remain valid, providing all other terms besides q and
3
ocn2 are disregarded even in the stratospheric accumulation zone.
4e have shown that ionic equilibrium conditions may also
be applied to the attachment of ions or radio-active atoms on natural
aerosol particles. Although their concentrations are extremely low,
compared with those of large and small natural ions, their experimental
study is already well in progress, in the neighborhood of the ground.
It presents an interest in the exchange layer where pollution and radio-
activity are still noticeable, but seems useless between the latter
and the tropopause. On the contrary, in the accumulation zone,
towards 20 km in height, where exist at the same time, radio-active
atoms (fission and artificial radio-activity) and an appreciable
quantity of pollution, it presents certainly an interest and could
constitute a new stage in the study of the atmospheric radio-activity.
pproved for Release: 2017/09/11 C06461858
Approved for Release': 2017/09/11 C06461858
REFERENCES
1. G.R.Wait and A.G. McNish - Month. Weath. Rev. 62
p�
1 1934
51.
2. P. Pluvinage et R. Utzmann - Ann. GeOphys. - 4, p. 161
1948
3- V.F. Hess - Ergebnisse der Kosmisch. Phys. 2, p.95 -
1933
4. H. '31'01 - Atmospharische Elektrizitat - Leipzig,
1957 und 1961
5. See '',rd report HASP- Vol. 2B
6. R. Mahleisen - Handb. d. Phys. - Vol. 48, p. 544, 1957
7- R.C. Sagalyn and G.A. Faucher - Journ. Atm. a. Terr.
Phys. 5. p. 253, 1954
8. L.B. Loeb - Basic Processes of Gaseous Electronics
University Press of California Berkley, 1960
9. E.H. Kennard - Kinetic theory of gases. new York and
Loudon, 1938
10. J. Bricard, J. Pradel, et A. Renoux - C.R. Acad. Sc.
Paris, 252, p. 2119, 1961
U. J. Bricard, J. Pradel, et A. Renoux - Geofisica pura
e applicata 51, p. 237, 1962
12. A. Renoux - to be published in C.R. Acad. Sc. Paris
13. J. Bricard -to be published in C.R. Acad. Sc. Paris
14. Rama and Honda - Journ. of geophys. Res. 66, p. 3227, 1961
15. 0.11. Gish and K.L. Sherman - Terr, Mag. 49, p. 159, 1944
16. H. Cauer - Archly Met. Geophys. Biokl. 1, p. 221, 1949
17. C. Junge - Journ. of Met. 11, p. 323, 1954
18. E. Evening und A. Wigand - Ann. Phys. 66, p. 261, 1921
19. C. Junge - Journ. of Met. 18, p..746, 1961
20. M. Deloncle - Journ. de Phys. 23, p. 269, 1962
R.G. Eldridge - J. Meteorology, 18, p. 171, 1961
S.
�rAt
!41
pproved for Release: 2017/09/11 C06461858
21.
22.
23.
24.
Approved for Release: 2017/09/11 C06461858
52.
J. Bricard, M. Deloncle et G. Tsranl - Ann. GoOphys.
15, p- 14t5, 1959
C. June - Der. Deutsch. Wetterdienst U.S.Zone No 35,
ID. 201, 1952
J.A. Chalmers - Atmospheric Electricity; Pergamon Press
1957
O. Thellier - Ann. Inst. Phys. Globe, Paris, 19, P. 107,
1941
25. R.C. Sagalyn and G.A. Faucher - Quart, Journ. Roy. Met.
Soc. . 62, p. 426, 1956
26. M. Smolukowsky - Z. Phys. Chem. 92,.p. 129, 1918
27. W. Holl - Beitr. z. Phy. d. Atm. 29, P. 83, 1956
28. Holl and R. Mnhleisen - GeofIsica pure e applicata
31, p. 115, 1955
29. J. Bricard, J. Pradel, et A. Renoux - C.R. Acad. Sc.
Paris, 253 - p. 1476, 196]
30. F.I. Scrase - Geophys. Mem. Met, of London, 140.64, 1935
31. w. D. Parkinson - Trans. of Oslo Meeting, 1948
32. J.J. Nolan and G.P. de Sachy - Proc. Roy. Ir. Acad.
". 37, 71, 1927
33. M.A. Fuchs - Iz - Akad - Nauk USSR - Geogr. Geophys
I - p. 341, 1947
34. J. Bricard - C.R. Acad. Sc. Paris, 226, p. 1536, 1948.
Journ. Geophys. Res. 54, p. 39, 1949
35. N.A. Fuchs - The Mechamis of Aerosols, U.S. Army,
Chemical Laboratories. Special Publications
CWL. 4, 12, 1955
36. D. Keefe and P.J. Nolan - Geofisica Pura et Applicata
50, p. 155, 1961
37. P. Pluvinage - Ann. de Geophys. 2, p. 31, 1946;
3, p. 2, 1941
38. R. Gunn - Journ. Met. 11, p. 339, 1954
39. M.H. Wilkening - Rev. Sc. Inst. 23, P. 13, 1952
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
53.
40. M. Kawano - Geofisica pura e applicata 51, p. 243, 1961
41. F.J.W. Whipple - Proc. Phys. Soc. London, 45, p. 367,
1933
43.
C. Junge - Journ. of Met. 12, p. 13, 1955
1sra81-renErf,
C. Ju2.Nuc1ear radiation in geophysics, p. 169
Springer, 1962
D. Keefe, P.J. Nolan and T.A. Rich - Proc. of the Roy.
Arlsb Acad. A - 60, p. 27, 1959
45. P. Arendt und H. Kallmann - Z. Phys. 35, p. 421, 1926
46. L. Lassen - Zeits.fbr Phys., 163, p. 363, 1961
47. J. Dricard - Geofisica pura e applicata, 51, p. 237,
1962
48. V.P.V. Flanagan et T.C. O'Connor. Geofisica pura e
applicata, 51, p. 145, 1961
49. J. Bricard. J. Pradel, et A. Renoux - to be published
in Annales de G6ophysique, 1962
50. H. Norinder - Geogr. Ann. Stockholm, I, p.1, 1921
51. F.J. Scraoe - Geophys. Mem. London, 67, 1935
52. M. Kawano Journ. of Geom. and Geoelectricity, 9,
D. 123, 1957
53. M. Kawano - Journ. of the Met. Soc. of Japan, 35,
p. 29, 1957
54. V.B. Milin i S.B. Malakev - Isv. Acad. Sc. U.R.S.S.
Section Geophy9. - 3, 1953
55. E.T. Pierce, L. Koenigsfeld, H. Isral, D.L. Harris -
in:.3.1itil:Rocent advances in atmospheric Electricity,
Pergamon Press, 1958
56. P.J. Nolan - Journ. Atm. and Terr. Phys. 6, p. 205, 1956
1.n1,1
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
11.14
����� -���� .aramm Avow. 24/1.4).��� ���������% �������������� moo.. ...��������� M.{ �����/....���������� ���� ������������ � ��������� M�MIIIV���� �01�1���������� �������������10 I
20 59 40
70./ figir_s/c3. sec
1-n-t-Lwe-44-4L4-ksottom
JornIs a 17'v 17
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
�r� so dr-. adh ma. s-4.� en�������������...� ���������....���mm.
e:b
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
islo -49.4/ Of �
- 3 ..... 0.49 -o6-30
N 4 - .15 -41-00
sca3 itw) N.+00/04)
�
21/22 jar1OCiry 19.54
fig. 4
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
pproved for Release: 2017/09/11 C06461858
�
1 1
OA 0,8
Approved for Release: 2017/09/11 C06461858
,1,ersurea1tn45 of/
X iiies-vt-�s-4, keefeirelem et Rich
Fortywiiii 23
at
II fornwlee 33
X Boltzmann
.TVN9t
0�01. ������
�mik IMO Mb M
�
�44444.4.
3 45
foson
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
210 aio- 4-.1045.10 10" Rorh-
en-c rn
ract 1 LIS
cril
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
tiT mi(N�) ALTITvog s.000 Pitts-- F i e t- .
t lit ore h'en I curve ( i'c n I 20,1--,�(211 12/ ecismic rays).
A its
r�c .-0-ws-4,-.14-44#i 9 ueS ) �,- 4
. .
� ' '4
thiorth'oP ci. i ne t. ( tc, i (I( On, i(vii'oo) . 4
8 � cau-pbe-i-ive-e-ri-ctu-e-04-rus-04K.wt I(l' le) A-:i
� -a.,
�.,..,
.:�A
ribroory
j 40 Ftivrter1953 x
Tr brag!), .
-
C.23147the-441troptett4erle i g ki-VP-}e-P1 953 �
( iliffmt
24 -
e t-peri'outtli-at Ctirve ft-M 1953 �
13 ticy 19 5st 0
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
f-4ferernritta( vet kiks
X Vairitrl- ex-pert-melt-Wes
ft-ertjetrerrayt
510-7 10" Z 3 510- 10'
rm.,/ S
in
C
15 . 9
pproved for Release: 2017/09/11 C06461858
f.;
Approved for Release: 2017/09/11 C06461858
SESSION 3.1.
Generation of Electric Charges Outside Thunderclouds
by
J. Alan Chalmers, M.A., Ph.D., P.Inst.P.,
Reader in Physics,
Durham Colleges in the University of Durham, England.
1. Introduction
It can scarcely be denied that the most important processes of electri-
fication in the lower atmosphere are those within thunderclouds, but never-
theless there are many other electrical processes at work in various regions
of the atmobphere that lam have been discovered and that will have small or
large effects on Use tl l it r ,t:t t
I,- t tY � u1t4. O. �I ft ..test.:;p1 le re..
It is hared that, at the risk or b.-cueing a intro.,: vutalogu�, numethIng
can be said about most of thene prooenees, though neternily there will be
more that can be discussed about sume than about others.
As in the thunderstorm, it is possible to distinguish setween two stages
in the generation of charge, first the actual separation of the two charges
of opposite sign from previously neutral matter and, second, the segregation
of these two charges in such a way that they reach different places. In
some of the pr insets discunned, the separtieion its that which produces the
naturnI cotehtetivIty or the Ionosphere, hrimeiy Ihr .1. atmoa-
phere by eonmie rnyn and by radit.ettivIty. Asal In 1,84.er111,e-n Ihr wgrrgatlan
Is by the ordinary process Cl' electric eonduction In the electrte field that In
present, moving charges of opposite sign in opposite directions.
2. Point Discharge
Point discharge Can be considered as a generator of charge-,,since the
breakdown separates charges from neutral matter and the electric field segre-
gates them.
When an earth-connected point exits above the uurface or the earth, the
lines of electric force in Use atmosphere concentrate on the poi:stand the local
.0
field strength tougdeater than that et the earth's surface. In the simple case
of an isolated point, this local field strength would depend on the potential
difference between the point and its surroundings, 1.e. the free atmosphere
pproved for Release: 2017/09/11 C06461858
�������
Approved for Release: 2017/09/11 C06461858
2.
at the same level, and so on the height of the point and the potential
gradient in the atmosphere; for a point which is one of a timber at eon-
parable heights, :he other points modify this simple approach.
When the potential gradiec.t in the atmosphere reaches a certain value,
the field strength near the point becomes sufficiently large for local
breakdown, involving loniratlon by collision, to occur, awl, aa the potential
gradient. inereunes, breakdown can occur over wider volumes and For more
points. The local breakdown p.r.s� points has seen ntadied under laboratory
conditions as "corona discharge", but the detailsrieed nut concern te: here.
The result of tics local breakdown is to produce Luna of both signs; those
of the same sign as the potential gradient move quickly into the point and
form a current to earth, while those of the opposite sign remain in the
atmosphere and Form a space charge, ultimately moving upwards to the cloud
whill Is the origin .1. Kradi,:ar... pre,wa,,....A. the npmee
eluirex teenr tip' .1.0i lit .1 lin int:dies the act out t.In�alo.s. al. 1.1.-
aa.ter nuitubIe eircsinstances the dis�ditirge may wear in pulses; .1.p,rag1ag
errveta ve:r perleds I. I. rug s�ogryvirt .1 with 1.11.� ietl;�te, it In ..ttey to
tier that, iii stendy ecesittionn, there will be ass teljustsnent to give it stend,y
current told space elairge, the local eentlitions smut the point providing just
sufficient current For the purpose; it is then unnecessary to discuss
details of the actual processes at work near :he point, just as, in an
analogous case of the Lanonuir-Child law for thermionic emission, the
details of the emission process need not be discussed.
With u particular point and on.: value of the potential grudlent,
follows that the current. would be altered only If the ntowl� ehttrifft 15
altered, toid thin Is achieved only by wind removing it Free the rmikhle�urhoed
of the point. An approxismte theorelleni calculation of the current In its
dependence on potential gradient and wind speed for an isoleted point has recently
been made (Chalmers, 1962) and work ls progressing on an attempt to carry out
a more accurate computation with a computer. The theoretical problem of a
point which is not isolated, but one of ft number similarly or differently
situated remains to tie ta,kled.
On the experimental side, the earlier measurements of the relation
between point-discharge current and potential gradient did not recognise
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
3.
the part played by wind, but the more recent work has shown that this must
be Included and the results are in fair agreement with the approximate
theoretical calculations.
The work so fur diseusned has been carried out by simply erecting a
metal point in the atmosphere and connecting it to earth through a measuring
� instrument; this, therefore, does not give a great deal of information about
the effect of natural point discharge, which must take place largely through
trees. The problem therefore arises as to how closely a metal point and a
tree correupond in regard LA point-divelatrge currents.
.Seltueltuni (1.98) out down a buttli, typical of the neighbourhood, mounted
It. on inn:dater:I and answered the current through it; while this wet; an
approach to natural condition:I, It did �not reproduce the conditions or a
living tree. More recently, Maenad and Chalmers (1960) uttempted to measure
�the current throudt a tree, not. directly but by measuring the effect of the
space charge on the potential gradient downwind; the results suggested that
a tree in leaf gives lean current than a point at the same height. Milner
end Clutlmers ( lent ) Inuorted 01ot:tr. 4.611 I 111.0 tt tree, :so nu to nhort-etreu II.
the eurrent. ihnsti the tree litrettelt a, gui vsouiou�tor foul round OW11 I.sr 'earnest.;
Until fur n point. quilm teteetiliy, with Ilse mew nipterntuu, Clottimern 0963)
a
found that, at. the thee tof.e.lose lightning flash, the tree rind a neighbouring
metal point gave appreciably different currents, showing that the tree does
not behave like a simple conductor. It would be very desirable if such
measurements could be attempted in the parts of the world where thunderstorms
are frequent.
The question of the effect of the presence or absence of neighbouring
point.e.on the current. through one purticetlar point la another which requirea
further itivetit teat Chiplenknr (1.9140) find Chelan:rat and tespitniOn (1955)
� tUntl potnt dinchnrge, round that. the total current, through is number.
of point.; clone Lowther woo lens than if they were replaced by ea single �
point, but Belin (191,8), in a laboratory experiment, found that each point
of a group gave the same current whether the others were present or not.
pproved for Release: 2017/09/.1-1 -606461858
t,g
-
V!
Approved for: Release: 2017/09/11 .006461858'
V:11
;41
4.
j. Non-Stormy Rain nnd Snow Cloud
Although the electrical effects in non-storew clouds are less than those
of thunderclouds, they are still appreciable, and statte processes of charge
separation must be present.
One of the important problems in this field is the discussion as to
whether the charge-separation processets in non-stormy clouds are the same as
In thundercloudu or not. If they are the mume procesues, then the problem
is why the magnitude of the charge separation is so different; what are the
conditions in the thundercloud that make the processes so much more efficient
there than in the non-stormy cloud? It should perhaps be pointed out that the
difference is much greater than the mere difference in intensity of precipita-
tion. If, on the other hand, charge separation in the non-stormy cloud takes
plaee by totally different processes, then these prureonee mat be ouch that they
. are not much magnified by the change from non-ntormy to atom, conditions, and,
further, the procenses in the thundercloud must be ouch that they cannot
operate significantly in the non-stoma, cloud. These considerations are a strong
justification for increased study of nonstormy cloud electrification, particularly
when it is realised that such clouds are much more frequent in msAY parts of the
world than thunderclouds and conditions are much steadier and more amenable to
measurement.
In this conneetion. tin important principle Ireta been used, nnmely that of
the quusi-uteady state, no that. one eats asnume that the total vertiral electric
current Is; the some at all levels. Thie asnumes that there Is no differential
horizontal electric current ut any level, and it might be profitable to consider
this question in more detail, particularly' in the case of warm-frout clouds
where the movement of the air is much more horizontal than vertical.
An important result in the measurements of the effects of non-stormy
clouds is the difference of both potential gradients and precipitation
eurrente fte between rtill11 anal C4010. Helene there ere effects at the earth's
surface, or near to It, lids 41,r1110 to ludlente that there meet be electrIcel.
effeetn lu the procenn or melting, stnee moist of the precipitntion concerned
hfts started as ice particles and, If finally falling an rata, has melted later.
The simple discussion of the quasi-steady state lends to the conclusion that
the potential gradient above a cloud would alter in sign when the precipitation
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
5.
changes from rain to snow or vice versa; it would be most desirable if this
conclusion could be confirmed or refuted by actual observations above the
clouds.
4. Non-Ruining Clouds
The charges in warm clouds which are not precipitating have been
measured by a number of workers and, although there is some disagreement,
the general results appear to be that the larger droplets more often carry
positive charges and the smaller negative, the actual charges being only a
few electronic units.
Some laboratory experiments by Barklic, Whitlock and Haberfield (1958)
have uhuwn that these chargeu are directly related to the presence of ions
and this 16 in accord with theoretical work by Gux, (1955).
Though these effects provide the only separation of charge in clouds
of this type, it seems unlikely that similar processes could be appreciable,
compared with others, in clouds which give precipitation.
Another effect in non-raining clouds is that which is termed the "traffic-
Jam effect". Since the conductivity within a cloud is less than that outside,
in order to maintain continuity of current acrosn a cloud boundary, there must
be a greater potential gradient within the cloud than outside, and this, In
turn, means a region of space charge at the boundary.
5. Precipitation
Precipitation, as it reaches the ground, is usually charged; if the
charge it carries is that which it has received in the cloud, then this is
not to be considered separately from the problems of charge generation within
the cloud; but if the charge on the precipitation has been altered as it falls,
then the processes by which this occurs must be considered as separate. If
precipitation is an important factor In charge generation In thunder clouds,
then there must be a large change of charge during full, since it is certain
that the precipitation current reaching the ground in much smaller than it
would be In the charge-generating region of the cloud.
When precipitation leaves the cloud, the only ways by which it would seem
possible that It could acquire charge would be 1) melting 2) capture of ions and
j) shattering.
Doroved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
6.
The fact that the charges and the potential gradients during snowfall
differ from those during rainfall (see, for example, Chalmers 1956), show
that it is likely that something occurs during melting and it is very
desirable that more should be found out about this, both by observations
during precipitation and, if possible, by laboratory measurements.
The "inverse-relation" between precipitation current and point-discharge
current (Simpson 19119) can be explained if the precipitation acquires charge
frum the point-discharge Ions and it more e-a._ t _e A I) a . connidcration of thlu
(CluLUners 19',1) has shown fairly satisfactory renuits. For the correlation of
rain current with potential gradient when there In no pulnt discharge, Inc
matter does not appuar so simple and further work is required. At the time
this is being written, work is in progress on the measurement of rain current
simultaneously at the top and foot of a 60-foot tower; results and conclusions
may be available by the time of the Conference.
Kelvin (l660) and Chauveau (1900) have found a change of sign of the
potential gradient at the top and foot or a tow!r on certain occasions during
rain, and this requires a negative npace charge in the air below the top of
the towrr; measurements of ehnre,eo doe to spinahIng Cannot entirely explain
throe resulto and it might be that shattering or mtin drops occurs in this
region.
6. The Electrode Effect
It is perhaps questionable unether the electrode effect should be
Included as a charge-generating process, since the actual charges concerned
are only the ions produced by cosmic rays and radioactivity. hut recent work
ha, euggeated that the electrode effect, Red theeconvection of the charge!,
separated by it, are of appreciable import/titre in the atmosphere.
The electrode effect cumprises a spare charge near an electrode, in
. the case of the atmosphere the earth's surface, and atises because there can
normally be no ions leaving the electrode, so that the conductivity at the
surface of the electrode is due to ions of one sign only. .In normal fine-
weather conditions, the electrode effect would give a positive space charge
near the earths surface; this is, in ruct, found only la special conditions,
e.g. over the Greenland tee-cop .(-tubnite, 1962) or over water (Mahleisen, 1961);
in other caues it is reduced or ablwnt hecaune of higher ionization close to
the earth.
A ooroved for Release. 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
7.
7. Other Natural Sources of Charge
There are several other natural phenomena which give rise to charges
in the atmosphere.
In blizzards, there are charges generated by, presumably, the impact of
6now particles on one another and on the snow on the ground (Simpson 1919) and
the Battl occurs in drifting snow. Obstacles such u3 wires become charged in
a blizz'ard (0arr4, 19',))�
Bust-storms give rise to quite large electric potential gradients,
sometimes even giving lightning (budge, 1914), and volcanos al DO give
,..,:�onsideruble effects to be ascribed to frictional effects of the ash
(liatekayama and Uchikawa, 1951).
. It has long been known that positive charge is generated by the
' splashing of writer, e.g. at waterfalls (Lenard, 1892); Mahleisen (1958) has
, .
()Una chargq:separation with change of humidity; there is evidence (mum elsen
,
A-1959) for positive charge originating at the sea-shore in the breaking of
wvea Blunehard (1)61) Isis r ound thn I, pus I t I vely Ovirged Igtr I it lov
, �
!ilOve upwards from the net nurrace, produced by thr breaking or ti 1r bubbles
the nes.
IL Is probable that only the laut two of these are likely to be of
appreciable .importance in the whole balance of charges In the atmosphere,
but there is scope. for more investigations of all these phenomena.
8. Artificial Sources of Charge
Mahleisen (195)) has invest igatvd hi detail the generation of charge
by burning and uther industrial processes and hau round that different
processes produce different signs of charge; the funountu of charge taken
into the atmosphere may be quite appreciable. An � earlier example of the
6MMC was the positive charge arising fru� locomotives (Kelvin, 113&)).
Chalmers, (1952) found that negative charges can be liberated into the
- atmosphere from high-tension cables in conditions of high humidity when the
insulation partially breaks down; sufficient charge is liberated to produce .
negative potential gradients several km. downwind�
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
8.
Another example of an artificial source of charge is the observation of
Moore, Vonnegut, Semonin, Bullock and Bradley (1962) who found positive space
charge downwind of a television tower in fine weather; this they explained
as due to the removal by wind of a space charge formed at the top of the
tower by the electrode effect.
Dell6erate attempts to produce space charges in the atmosphere by
electric discnarge have been made by Vonnegut and Moore (1958), Vonnegut,
Maynard, Sykes, and oore (1961) and Vonnegut, Moore, Stout, Staggs, Bullock
and Bradley (1962).
References
Barklie, HOLD., Whit luck, W. and Haberflold, G. 1958, "Ona.,rvutJon�
on the reactions between small lona and (a) cloud droplets, (b)
Aitken nuclei", in "Recent Advances in Atmospheric Electricity -
Proceedings of the 2nd Wentworth Conference, May 1958" (ed. L.G. Smith,
New York: Pergamon Presn).
Barre, M. 1953, "Propr14te's electriques du blizzard" Ann. G6ophys. 2,
pp 164-183.
Belin, R.E. 1948, "A radio-sonde method for atmospheric potential gradient
measurements", Proc. Phys. Soc. Lend. 60, pp 581-5B7.
Blanchard, D.C. 1961, "line ,AeetrIfIcntion of the atmosphere by partLcieu
from bubbhnn in lhe sea", Unpublished paper from We, Ride Oceanographic
Institution (ix:fel-emu 61-9).
Chalmers, J.A., 1951, "The origin of the electric charge on rain", Quart.
J.R. Met. Soc. fl, pp 249-259.
Chalmers, J.A., 1952, "Negative electric fields in mist and fog",
J. Atmosph. Terr. Phys. 2, pp 155-159.
Chalmers, J.A., 1956, "The vertical electric current during continuous rain
and 6110e, J. ALmouph. Tcrr. Phys. 2, pp 511-521.
Chalmers, J.A., 1965, "Point-discharge currelits through ft living tree during
ft thunderstorm", J. A9nuspin Terr. Phys.
Chalmers, J.A. and Mapleson, W.W. l95!5, "Point discharge currents fran a
captive balloon", J. Atmosph. Terr. Phys. 6, pp 149-159.
Chauveau, B. 1900, "Etudes de la variation de l'electricite atmospherique",
Annales de BCM 5, p.l.
pproved for Release: 2017/09/11 C06461858
JWV
Approved for Release: 2017/09/11 C06461858
9.
Chiplonkar, M.N., 1940, "Measurement of point discharge current during
disturbed weather at Colabu", Proc. Indian Acad. Sci. A 12, p.50.
Gunn, R. 1955, "The statistical electrification of aerosols by ionic
diffusion", J. Coll. Sei. 10, pp 107 - 119.
Hatekayama, H. and Uchikawu, K. 1951, "On the disturba:,ce of the atmospheric
potential gradient caused by the eruption smoke of he volcano Aso",
Gen. Ass, Int. On. Geod. (Brussels), Ass. Terr. Mugn. Elect.
Kelvin, Lord, 1860, "Atmospheric Electricity", Royal Institution Lecture;
reprinted in "Papers on Electricity and Magnetism" (London : Macmillan & Co.)
pp. 208-226.
Lenard, P. 1892, "rber der Elestrizitrit der Wasserfrille", Ann. Phys. Lpz.,
46, pp. 584-636.
Maund, J.E. and Chalmers. J.A. 19o0, "Point discharge from natural and
artificial points", Quart. J. R. Met. Soc. 86, pp. 85-90.
Milner, J.W. and Chalmers, J.A. 1961, "Point discharge from natural and
artificial points (Pt II)," Quart. J. R. Met. Soc. 87, PP. 592-596.
Moore, C.B., Vonnegut, B., Scwonin, R.G., Bullock, J.W. and Bradley, W.,
1961', "Pair-weather atinuspheric ,�.[Cl' LI' I(' pol.ruf 1/11 gri I c-Ilt it lid illown
charge over Central Illinois, Stumner 1960", J. Geophys. Run. 67, pp.
1061-1071.
Mftleisen, R. 1953, "Die luftelektrischen Elemente in GrossstadtberSich",
Z. Geophys. 29, pp. 142-160.
Matileisen, R., 1958, "Elektrische Ladungen mu! Kondensationskernen bei der
Wasseraufnahme und - abgabe", Miturvissenschaften 45, pp 1-2.
Mahlelaan, R. 1959, "Die luftelektrlsch.,n 10,r1Mtnisse im KUstenaurosol, I",
Arch. Met.. Wien A ll, pp. 95-108.
MMUleisen, R. 1961, "Electrode effect measurements above the sea", J. Atmouph.
Terr. Phyo. 20, pp 79-81.
Rude, V.A.!)., 1915, "On the electrification associated with dust clouds,"
Phil. M2g. 22, pp 481-494.
Ruhnke, L.R., 1962, "Electrical conductivity of air on the Greenland ice-cap",
J. Geophys. Res. �7, pp 2767-2772.
Schonland, B.F.J. 1928, "The interchange of electricity between thunderclouds
and the earth", Proc. Roy. Soc. A. 118, pp 252-262.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
10.
Simpson, G.C. 1919, "British Antarctic Expedition 1910-13 Meteorology",
(Calcutta), 1, pp 302-312.
Simpson, G.C., 1949, "Atmospheric electricity during disturbed weather",
Geophys. Mem., Lond. 84, pp 1-51.
Vonnegut, B., Maynard, K., Sykes, W.G. and Moore, C.B., 1961, "Technique
for introducing low density space charge into the atmosphere",
J. Geophys. Res. 66, pp 823-830.
Vonnegut, B. and Moore, C.B., 1958, "Preliminary attempts to influence
convective electrification in cumulus clouds by the introduction of
space charge into the lower atmosphere" in "Recent Advances in Atmospheric
Electricity - Proceedings of the 2nd Wentworth Conference, May 1958"
(ed. L.G. Smith, New York: Pergamon Press) pp. 317- 331.
Vonnegut, B, Moore, C.B., Stout, G.E., Staggs, D.W., Bullock. J.W. and
Bradley, W.E., 1962, "Artificial modificatbn of atmospheric space charge",
J. Geophys. Res. _6_1, pp 1073-1083.
Approved for Release: 2017/09/11 C06461858
;
Approved for Release: 2017/09/11 C06461858
CHARGE GENERATION IN THUNDERSTORMS
By B.J. Mason
(Imperial College, London)
Introduction
It is proposed that the principal mechanism of thunderstorm
electrification involves the accretion, freezing and sFlintering of
supercooled droplets on pellets of soft hail. Gravitational separation
of the small positively-charged ice splinters and the much heavier
negatively-charged hail pellets then produces an electric field of
the observed polarity.
Evidence in support of this theory comes from: (1), observation
of the disposition of electric charges and fields in thunderclouds;
(ii), observed correlations between the appearance of soft hail and
strong electric fields; (iii), laboratory observations that riaing
elements acquire a negative charge as positively-charged splinter*
are ejected from freezing drops; (iv), the discovery that this
separation of charge arises from a basic property of ice, viz �
protonic thermo-electric effect which has been investigated experiment-
ally and theoretically in some detail; (v), application of the
laboratory results on the rate of charging of artificial hail pellets
to a model thunderstorm which reveals that the proposed mechanism is
capable of producing and separating charge at the rate required by
observations on lightning flashes while other mechanisms appear to
work much too slowly. These arguments will now be presented in
more detail, but first it seems worthwhile to list the more important
and relatively undisputed features of the thunderstorm with which any
satisfactory theory of electrification must be consistent.
2. Requirements of a satisfactory theory of thunderstorm electrification
The theory must explain quantitatively how electric charge is
generated and separated in a thundercloud at a rate equivalent to that
at which it is dissipated in lightning flashes. It must account for
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-2-
the observed polarity of the thunderstorm and be consietent with what
is known about the electrical and dynamical structure of the storm,
the nature of the electrical field changes accompanying lightning, the
size and duration of the storm, and the nature, scale and intensity of
the precipitation processes which are generally considered to be
closely correlated with the electrical activity. More specifically,
the theory must be consistent with the following facts:
(i) The average duration of precipitation and lightning from a
typical single thunderstorm cell is about 30 min.
(ii) The average electric moment destroyed in a lightning flash is
about 100 C.km, the corresponding charge being 20-30 C. A typical
cell produces flashes at intervals of about 20 sec 80 the average
lightring current is about 1 amp.
(iii) The magnitude of the charge which le being separated immediately
after a flash, by virtue of the falling speed of the precipitation
elements, is of order 1,000C.
(iv) 41 a typical cell this charge is generated and separated in a
volume bounded by the 0 and -40�C levels and having a typical radius of
2 km and therefore a volume of about 50 1cm3.
(v) The negative charge is centred near the -5�C level, while the
main positive charge is situated some kilometree higher up; a
subsidiary positive charge often exist near cloud base where the
temperature is usually a little warmer than 0�C.
(v.0 Sufficient charge must be generated and separated to supply the
first lightning flash within 10-20 min, of the first appearance of
precipitation particles large enough to produce a radar echo.
In round figures, the requirement is to generate about 1000 C of
charge in a volume Of about 50 km3 in a period of about 20 min. i.e. at
an average rate of 1 C/km3/min.
3. Observational evidence for an ice mechanism
(i) Lightning is usually accompanied by heavy precipitation although,
1
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-3-
in warm, dry climates, thin may not reach the ground. Most theories
on the origin of the electric charge have assumed that the precipitation
plays an important role, and as the main charge centres appear at
levels in the cloud where the temperature is below 0�C, it is natural
to associate that generation with the presence of supercooled water
and/or the ice phase.
(ii) Kuettner (1950), from observations made inside thunderclouds
capping the Zugapitze in Germany, reported that solid precipitation
elements were predominant in the greater part of the thundercloud and
were present on 93% of the occasions. Snow pellets and pellets of
soft hail were the most frequent form of hydrometeor being present on
75* of occasions but large hail was relatively rare.
(iii) Fitzgerald and Byers (1962), using aircraft fitted with
electric-field meters, have reported that the actively building regions
of thunderstorms are regions of excess negative charge. The strongest
fields, of up to 2300 Vies, were associated with regions of heavy
precipitation. In particular, a large hail shaft produced a strong,
smoothly increasing field indicating a negative charge on the hall.
(iv) Malan and Schonland (1951, a, b) find that, in South African
storms, the negative charge its often distributed in a nearly vertical
column which may extend up to but not beyond the -40�C level. Thie
is consistent with the charge being generated by growing hail pellets
because supercooled droplets exist at temperatures down to, but not
below -40�C.
4. The charging of rime deposits
In recent years, several workers have reported that when super-
cooled water droplets impinge and freeze on an ice surface, the
resulting layer of rime acquires a substantial charge. The experimental
results, which have been reviewed by Mason (1957), may be summarized
as follows.
Findeisen (1940, 1943) formed a rimed layer by spraying water
droplets on to a cold metal surface and found that it acquired a
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-4-
positive charge. The charging ceased if the surface became smooth
and glassy, as was the case if the drops froze slowly, or if it
became wet. Rather stronger charging was obtained with a natural
supercooled cloud than with an artificial spray, the difference being
ascribed to more rapid freezing of the smaller cloud droplets. The
rate of charging in the cloud was 3 x 10-13 C c12 s-1.
In a later investigation, Kramer (1948) found the rime deposit
acquired a negative charge which increased in proportion to the
impact velocity of the droplets. With a velocity of 0.5 m a-1
the charging rate was 2 x 10-14 C cm-2 s-1 and, for a velocity of
5 m-1, ten times larger.
Lueder (1951 a,b) made experiments in natural supercooled clouds
on a mountain top in order that the coLtaminants in the water should
be those occurring in nature. He states that the growing rime deposit
acquired a negative charged, an equal positive charge being communicate
to the air, probably on the parts of the drops which were flung off
without freezing. Unfortunately, it is difficult to interpret hie
experiments and to deduce the actual rate of charging.
Meinhold (1951) measured the electric field strength at the
surface of the fuselage of an aircraft flying at 80 m e-1 through a
supercooled cumulus congeetus cloud. The deposition of rime was
accompanied by � rapid rise in the field strength in a sense which
indicated that the aircraft was acquiring a negative charge, and the
rate of charging was calculated to be 5 x 10-12 C cm-2 a-1.
The charging of a rime deposit on a cold metal surface was also
studied by Weickmann & aufm Kampe (1950). Water droplets in the
diameter range 5 to 10G!6 were sprayed at velocities varying from
5 to 15 m a-1 on to a metal rod of 5 mm diameter in a cold room kept at
either -5 or -12�C; they were therefore slightly supercooled on
reaching the rod. The rate of charging, which was not sensitive to
the presence of dissolved salts, increased with increasing velocity of
-1
the air stream, and for a velocity of 15 � a , attained a value of
1
--Approved for Release: 2017/09/11 C06461858--
Approved for Release: 2017/09/11 C06461858
-5-
1
5 x 10-12 -
G cm2 When water at temieraturea slightly above
0�C was sprayed on to the rod it acquired a slight positive charge.
Later the authors indicated that the results of these experiments
may have been seriously affected by electrification associated with
the production of the spray.
The balance of the evidence from all these experiments points
to the acquisition of a negative charge by a growing layer of rime,
Findeisen's result being an outstanding contradiction. In light
of the recent experiments of 1,,,tham and Mason (1961), described below,
it now appears that some of the differences between the results of
different workers may be ascribed to the use of differing drop sizes,
temperatures, and impact velociyes, while epurious effects may arise
from initial charging of the spray droplets and electrification
produced by the splashing of droplets on the ice surface.
5. The splintering and electrification of freezinK water drops.
Some evidence for the production of charged splinters by a
growing rime deposit was obtained by Kramer (1948), and their
production during the freezing of individual water drops was
investigated in detail by Mason and Maybank (1960).
Nucleation of a water drop, at temperature -T�C, is followed
by rapid solidification of a fraction T/80 of its mass in the form
of an ice shell. Subse4uent freezing of the liquid interior now
proceeds at a rate determined by the disnipation of the latent heat
to the surroundings. The expansion which accompanies this freezing
sets up stresses in the ice shell which may disintegrate to produce
a number of ice splinters. Mason and Maybank found that, for drops
suspended in still air, the number of splinters produced was almost
independent of the drop diameter in the range 0.1 to 2 mm but that
for drops of diameter < 60I., srlinter production was much reduced.
They also measured the charges on the residues of fragmenting drops.
If only a minor fraction of the drop was blown off the residue was
invariably negatively charged. Typically, a drop of 1 mm diameter
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
�6�
freezing at -50C produced about 20 splinter and acquired a
negative charge of about 10-4 e.s.u.; a similar drop freezing at
-15oC produced about 5 splinters and acquired a charge of about
3 x 10-5 e.s.u. The average positive charge per splinter was
therefore 5 x 10-6 e.s.u. These experiments suggested an explanation
for the electrification of growing rime deposits and that a similar
mechanism operating during the growth of hail pellets might be an
important factor in the electrification and ice-crystal economy of
clouds.
6. Charging associated with the growth of soft hail pellets.
Latham and Mason (1961) measured the electrification of artificial
pellets of soft hail as they grew by the accretion of supercooled water
droplets, and determined how thia varied with the temperature, size
and impact velocity of the drops.
The experiments were conducted in a cold room, at air temperatures
ranging from 0�C to -17�C, with the apparatus shown in Fig. 1.
The hailstone, simulated by a 5 mm diameter, electrically
insulated, copper sphere coated with tk a com layer of ice, was
suspended in the centre of an earthed vertical brass tube through
which the air stream carrying the droplets could be drawn at velocities
ranging from 0 to 30 m a-1. Water drops of uniform diameter in the
range AO to 90ja produced by the spinning-top apparatus of Walton &
Prewett (1949), or rather larger drops produced by an atomizer, were
allowed to fall several feet in the cold room where they became super-
cooled to very near the air temperature before reaching the hailstone
target. For a given droplet size and air-stream velocity, the flux
of droplets hitting the target was determined by allowing them to strike,
for a given time, a Formvar-coated glass sphere of the same dimensions,
and countihg the droplet impresaions under the microscope. The
impaction and freezing of the droplets was accompanied by the ejection
of ice particles from the target aurface; their number and sizes were
determined by inserting Formvar-coated alidea just beneath the hailstone
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-7-
and later examining the plastic replicas of the crystals.
The electric charge accumulating on the hail pellet during the
freezing of droplets on its surface waa measured by a Vibron
vibrating-reed electrometer of resistance 101211. and time constant
200 e. The minimum detectable charge was about 5 x 10-4 e.s.u.
A
5
2
-Ot
The surface temperature of the target, measured by a thermo-
couple, was higher than that of the surrounding air because of the
latent heat released by the freezing drops and could be raised
artificially by irradiating the surface with the beam of a 50 W
tungsten lump.
The experimental operations, which could be performed from
outside the cold room, consisted of setting the cold room (air)
temperature, the air-stream velocity and the drop size and reading
the electrometer after the target had been exposed to the droplets
fora known time, usually 10 e. Then the effect, on the rate of
charging, of raising the surface temperature of the hailstone (this
being previously calibrated in terms of the air ispeed and the current
etipplied to the lamp) yea inveatigated. Meanwhile, slides for
collecting the ice crystals shed by the target were ineerted at
regular intervals. The whole procedure was then repeated for a
different set of conditions.
The freezing of droplets of distilled water on the surface of
the hailstone caused it to become negatively charged and WSS SCCOM-
penied by the ejection of small ice splinters. The manner in which
the average charge and number of aplintere produced per drop varied
with the drop diameter, imloct velocity and air temperature is shown
in figures 2, 3 and 4_
In a typical experiment, with the air temperature at -15�C and
4
the air etream moving at 10 m a-1, 10 drops of diameter 80/, struck
the hailstone within 10 e and produced a total charge of 4 x 10-2 e.s.u.,
i.e. an average charge of 4 x 10-6 e.s.u. per drop. On average, each
droplet produced 12 ice splinters; their mean diameter on collection
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
f.�
-8-
was 201s but they were probably smaller on ejection and had grown
in the meantime.
Droplets of d< 301/6, produced few splinters and little charging.
Figure 2 shows that the production of both was enhanced as the droplet
diameter was increased to about 50r , remained fairly constant for
diametera between 50 and 80i-, and fell again for still larger
drops. These results are in fairly good agreement with those
which Mason and Maybank (1960) obtained for individual droplets
suspended on fibres except that they did not observe a reduction in
splintering for large drops. Thin tendency, in the present experi-
ments, may be exclained by the fact that, in impinging at several
metres per second, the larger drops shattered before freezing and
that (*lashing communicated a positive charge to the ice target in
the manner observed originally by Faraday and more recently by Gill
& Alfrey (1952). Positive charging of the target at high impact
velocities is shown in figure 3; this occurred although there was
still a considerable production of splinters.
An shown in figure 4, the rates of charge and splinter production
are almost independent of the air temperature in the range -6 to -17�C
but both fall off rapidly at higher temperatures and, in our experimenta
were no longer detectable at -2�C. The explanation is as follows.
The impacting droplets can be frozen only at a rate determined by
the rate of dissipation of their latent heat to the environment; at
air temperatures close to 0�C the rate of freezing was slow and
consequently the hailstone surface became wet and, as the replicas
showed, considerable aplashing occurred as the drops struck it.
A number of tests were carried out to make sure that charging of
the hailstone was due entirely to the collision and freezing of the
droplets. No detectable charging occurred when the air stream
carried no droplets and when droplets, impacting at very low velocity,
froze on the surface without producing splinters. The parallelism
tit
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-9-
between the curvea of charge and aplinter production in figures
to 4 is, perhaps, the strongest evidence for the one being a
consequence of the other.
Irradiation of the hailstone by the lamp caused a reduction in
the rate of charging. For example, with the air temperature at -12�C,
raising the surface temperature of the hailstone by d�C reduced the
rate of charging by about 20%; when the surface was warmed by 5�C,
the charge production was halved; but, in both cases, the rate of
splinter production was not appreciably altered.
When the impinging water droplets were contaminated with sodium
chloride in concentration corresponding to the average found in cloud
water (3.6 mg/1.), the rate of charging was decreased by about 20%.
These experiments appear conclusive in showing that the negative
charging of the hailhtone is caused by the ejection of small splintera
of ice during the freezing of droplets on its aurface. The influence
of droplet size and the presence of salt are in fair agreement with
the observations of Mason and Maybank apart from the effects which
were produced by splashing of large drops.
In a later paper, Latham and Mason (1962) reported that when the
above experiments were repeated in the presence of an external electric
field the charging rates of the hailstones were altered by only about
10 per cent by application of fields of-. 1000 V cm-1. The inference
is that the charging of hailstones will not be greatly accelerated by
the cumulative build-up of polarizing fields in thunderatorms.
7. Application of laboratory results to computation of the
production of charges and electric fields in a model thundercloud.
We consider a thunderstorm in which above a level Io the
updraught U contains an exponential size epectrum of hailstones such
that the concentration of stones within the radius interval R to R dR is
where No and-A. are constants indepdndent of position in the updraught.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-10-
Since splinters and charge are produced only by the freezing
of droplets on the surfaces of hailstones which remain Da, we write
the fractional volume FD(z) swept out per unit time by the dry hail
at level z as
I y(t)z) H(z) tR
(2)
where V is the hailatone fallspeed and .is the critical radius
at which a hailstone becomes %.et at level z. We now aaaume that this
flux of dry hail encounters a concentration n(z) of supercooled drop-
lets of radii greater than 251..; smaller dropleto.produce few splinters
and little charge. Since these droplets are continually being swept
up by the entire spectrum of hailstones, their concentration will
decrease with increasing height according to
e-p H7:
.tz) )
(= )
(3)
where no is the concentration of dropu at z and F(z) refers to the
0
entire hail spectra.
The total rate of charge production between levels zo and z owing
to the impaction, freezing and splintering of large cloud droplets on
dry hail is then given by
Z.
cl6) A 41,, IA. F, (7) 241) P--(Z) etz) cLz
where A is the mean cross-sectional area of the updraught and is
the average charge produced by the freezing of a drop of ri 25/.-
We have seen that a typical single-cell worm generates about
1000 coulombs of charge in about 20 min., i.e. at an average rate of
about I amp. Putting = 1 amp = 3 x 109 e.a.u., A = 91r x 101� .92
( a mean radius of 3km for the cell), qd = 4 x 10-6 e.s.u. (Latham
and Mason 1961), and values of FD and F based on the observations of
Atlas and Ludlam (1961) for a storm having an updraught increasing
linearly with height according to U(z) a 5(z-1) m sec-1 with z
measured in km and a liquid water content of 3 gm-3, we find no = 5 cm-3.
_Approved for Release: 2017/09/11 C06461858-- -
Approved for Release: 2017/09/11 C06461858
-11-
This does not seem an unreasonable value since droplets of r > 25,/,
have been found in concentrations exceeding 1 cm-3 in rather small
cumulus (Durbin 1956), while concentrations of 5 cm-3 in Cb have
been observed by Weickmann and aufm Kampe (1953).
The corresponding concentration of splinters, assuming a freezing
drop to produce on average, 10 splinters, is calculated to be
about 1 cm-3 between the -20�C and -50�C levels (Browning and Mason,
1963). This, too, seems a reasonable figure.
The rate of building of the vertical electric field E by separation
of the hail pellets and ice splinters is given by
4 /3 etlz Tr ircR) (1,2 v*60 (f)
oa;
311 (14J(V/k)
where re represents the leakage of charge through point-discharge
and conduction currents, p the precipitation intensity, j7,.. the
average density of the ice i_articles, V and *II are respectively
weighted mean values for the fall speed and mean radius of the hail-
stones. For particles sizes ranging from R 0.1 mm to 1.0 mm,
ciral x const 8500. If we assume that the precipitation
intensity increases rather rapidly with time at first, and then
levels off at a value pm.which is maintained for several minutea,we
may write ID r�, - and Eq 5(a) becomee
where
gLIE 4 71 ore
ct C
- 3Ir x bscrt) (41,15!2.9 p..
The solution of (6) with the conditions E = 0 and d4/41.- = 0
when t = 0 is
cr 1 - -P/3 4- )
((.3 -.4 0, Fa
Taking , . 2 x 10-3 e,a.u. and A . 5 cm hr-1, i?.. . 0.5 g cm-3
_L m 600A 1 nk. = 1 cm- * 3 and 1, : 4 x 10-6 e.s.0 Eq (6) predicts
.4 ,
that the field will reach a value of 9400 V cm-1 within 10 min of the
appearance of precipitation elements, but, before larme-scale
fields of this magnitude are reached, lightning dischargea will almost
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
t.f
ii
-12-
certainly occur and destroy the field. However, it appears that
the accretion, freezing and splintering of supercooled droplets on
hail pellets can readily account both for the production of charge
in thunderstorms and for the generation of large-scale fields of
several thousand volts per cm during periods of about 10 min.
If the field were destroyed by a lightning flash but the soft
hail continued to fall at a steady rate p� , the field would
recover at a rate given by
001-7
de,c1-
or
,r
) +/: e:/3(7
/5 (32 /3
(7)
(?)
With /3 and cr taking the same values as above, the field would
build up again to 4000 V cm-1 in 30 s, which is about the average
interval between lightning flashes from a modest single-cell storm.
t The basic mechanism of charge-separation during the freezing and
splintering of droplets.
The fundamental problem is to eiqlain how the electric charge can
be separated during the bursting of a freezing drop and why the
splinters are ejected with 2 positive charge leaving a negative
charge on the hailstone. Mason and Latham (1961 ) have troposed
that this is a munifehtation of a protonic thermo-electric effect
in ice by which the hydrogen and hydroxyl ions formed by the
dissociation of a small fraction of the ice molecules become separated
under the influence of a temperature gradient.
The rrocess depends essentially on two facts. One is that the
concentrations of positive and negative ions increase quite rapidly
with increasing temperature; the other, that the hydrogen ion (proton)
diffuses such more rapidly through the ice crystal than does the
hydroxyl ion (Eigen and de Maeyer, 1958). Thus, if we imagine a
steady temperature difference maintained across a liece of ice, the
warmer end will initially possess higher concentrations of both
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-13-
positive and negative ions. The more rapid diffusion of H' ions
down this concentration gradient leads to a separation of charge,
with a net excess of positive charge in the colder part of the ice.
A detailed theoretical treatment by Mason 4sa=bat*Ian7(111.61) leads
to the following expression for the potential V produced by a
temperature difference T across the ends of an ice specimen
i) (,
= irG
millivolts
(cj)
where U , U are res!_ectively the mobilities of the 1.1+ and OH- ions.
14/0_= 10, e the electronic charge, k Boltzmann's constant,
15 . 1.2eV the actiwition energy for dissociation in ice and T is
the absolute temperature. Latham and Mason (1961) have measured the
potential differences across specimens of pure ice and find excellent
agreement with Eq. (9). Moreover, the potentials are not markedly
affected by the presence of dissolved salts and gases in the ice.
When two pieces of ice at different temperatures are brought into
temporary contact and separated, the warmer piece acquires a negative
charge and the colder one an equal positive charge . Theory indicates
that there should be a maximum charge separation of 3 x 10-36 T e.a.u.
between each cm2 of contacting surface when the surfaces are separated
after about 0.01 sec. If they are left in contact for longer times,
the charge separation will be decreased as the two pieces of ice become
more nearly equal in temperature. These conclusions have been
confirmed by experimental measurements that show that very little charge
separation occurs if the contact period exceeds 0.5 sec.
The electrificnticn of freezing water droplets is explained by the
preferential migration of protons down the temperature gradient
established across the ice shell. During the early stages of freezing,
a radial temperature gradient is established across the ice shell, the
inner surface being held at 00C by the water that is still liquid inside,
the outer surface cooling towards the air temperature. According to
the above theory, protons will migrate preferentially down this
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-14-
temperature gradient and produce an excess poeitive space charge
in the outer layers of the ice. When the droplet bursts by the
expansion of the centre as it freezes, aplinters ejected from
the outer layer will tend to carry away a positive charge and leave
the remainder of the drop negatively charged. This is in accord
with the experimental observations.
It is difficult to calculate the temperature gradient across the
ice shell, which varies with time, and therefore the space charge
density in the outer layers at the time of rupture. During the
formation of the initial shell, the temperature of the drop ia
everywhere 0�C; as freezing of the interior proceeds, a temperature
gradient is built up; when freezing is completed the droplet finally
assumes the air temperature everywhere and again the temperature
gradient diaappears. However, near the end of the freezing process,
when the centre of the drop is still 0�C and the temperature of the
outer surface close to that of the air, we may take the average
temperature gradient across the ice to be
T./q.
In fact, shattering
usually occurs before this late stage but we may use this value of
the gradient to calculate a minimal value for the separ,ted charge.
according to Eq. (9), this will be 4Are't. S"./0
e,s.u. If
T = -15�C and r 40/- , thin bac�mes 4 x I0-5 e.s.u. If r 0.5 mm,
a
the charge becomes 5 x 10-4 e.s.u. rhus the measured charges of
5 x 10-6 e.s.u. and 3 x 10-5 e.s.u. could be accounted for by the
fragmentation of one-tenth of the surface area of the dropa,
The Reynolds-Brook Charging Mechanism
Reynolds (1954), Reynolds, Brook and Gourley (1957), have
attributed the negative charging of hailstones not to the freezing
and splintering of droplets but to collisione between the hailstones
and much cmaller ice cryatals. Laboratory experimenta in which an
Ice sphere was rotated in a mixed cloud of ice crystals and super-
cooled droplets showed that the spLere acquired a negative charge;
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-15-
but if the cloud wan comlobed entirely of droplets or entirely of
crystals, there was negligible charging.
Reynolds believed that charging was caused by rubbing contact
between the simulated hail pellet and the ice crystals which bounced
off with a positive charge, and that the sole function of the
freezing droplets was to warm the rimed surface and to create
a temperature difference between it and the colliding crystals.
It was estminated that, with a temperature difference of a few
degrees, the average charge carried away by a crystal of radius
50t, was 5 x 10
Following up these ideas, Reynolds, Brook and Gourley (1957)
investigated the electrification which resulted from rubbing contact
between two rods of ice of different temperatures. When both pieces
of ice were formed from distilled water, their resistivity being about
108 Jl.cm, the warmer became neg:tive after rubbing. If, however,
one of the ice specimens was made from a 10-4 molar eolution of NaCl
it became negative even though it was 25�C colder than the 'pure' ice.
This reversal of potential was attributed to the formation of a liquid
layer during rubbing and, on refreezing, a selective incorporation of
Cl- ions into the colder ice. Later, Brook (1958) investigated the
potentials developed when two pieces of ice of different temreraturee
are brought into temporary contact and separated with a minimum of
frictional contact. The sign of the charging was related to that
of the temperature gradient, for both 'pure' and 'salty' ice, in the
same manner as before, but the potentials were an order of magnitude
smaller than those developed during rubbing contact.
These phenomena have been re-investigated by Latham and Mason
(1961, a, b). They found, in agreement with Reynolds et al, that
when two pieceb of highly purified ice were brought into temporary
contact (with minimum friction) and separated, the wormer acquired
a negative charge. A theoretical calculation bused on the protonic
thermo-electric effect, indicated that a maximum charge transfer of
3 x 10-3A T e.s.u. cm-2 should occur with a contact time of about
1
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858 -
-16-
1/100 sec. and that it should thereafter decline as the two
pieces become more nearly equal in temperature. The theoretical
value for the charge developed for a contact time of� 1/100 sec.
was well confirmed by experiments which also showed that very
little charge separation occurred if the contact period exceeded
4 sec. Contamination of the ice with 002, HF, and NaC1 in
concentrations of up to 50 times that normally present in rain-
water, did not greatly influence the electrification.
Latham and Mason also investigated the charging of a simulated
hailstone by colliaions with ice crystals. In order to eliminate
charging by the freezing of aupercooled droplets, these were
rigidly excluded from the cloud, and the surface temperature of the
hailetone Man controlled by an internal electric heater or by
irradiating its surface. Charging of the hailstone by the
colliding crystals was measured as a function of their tea;perature
difierence, and of the size and impact velocity of the crystals.
The sign of the charging was directly proportional to the temperature
difference but rather insensitive to the size (diameter ranging from
?.0/6 to 50/..) and impact velocity (1 to 50 m bec-1) of the cryetals.
with a temperature differ,nce of 5�C, it reboundinv, crystal of diameter
50/A produced, on average, a charge of 5 x 10-9 e.s.u.
This is five orders of magnitude smaller than Reynold's value
of 5 x 10-4 e.s.u.:
Reynolds' value may be questioned on two grounds. First, the
I
field produced by such a charge at the crystal surface would exceed
10,000 V cm-1; surely a discharge would occur between the pointed
crystal and the hail pellet before the charge on the crystal could
build up to this value.
Secondly, the charge of 5 x 10 e.s.u. is about 1000 times
greater than that of all the charged carriers that would be present
in the ice crystal if this were pure ices Additional carriers
might be produced as the result of local fricticnal heating of the
ice at the points of contact but even if the whole ice crystal
pproved for Release. 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
-19-
fall in a vertical electric field and collide with the much smaller
cloud-droplets that they overtake. The raindrop will be electrically
polarized by the field; in a down4ardly directed field such as
exists in the atmosphere in fine weather, the lower half of the drop
will carry a positive charge and the upper half an equal negative
charge. Elster and Geitel sugl:ested that, after collision, some
of the cloud particles would rebound from the underside of the rain-
drop, carry away some of its poaitive charge, and leave the raindrop
with a net negative charge. The falling drops, in carrying their
negative charge towards the base of the cloud, would enhance the
original field, and no the whole electrification process would build
up rapidly. A detailed mathematical treatment of the problem ahowa
that if only one per cent of the cloud droplets striking the lower
half of the raindrops were to bounce off, and if this percentage
were independent of the electric field strerirth, the field would
grow exponentially with time and, in a cloud producing ruin at the
rate of 1 cm/h, would reach 1000 timea ita initial fine-weather value
In only 10 minutes.
Now laboratory experiments indicate that, in the alv.ence of an
electric field, the fraction of cloud droplets that actually rebound
after striking raindrops is small; the great majority of collisiOns
result in coalescence. However, no experiment sufficiently accurate
to detect non-coalescence to the extent of only a few per cent has
yet been performed. This problem, which becomes comilicated further
by the presence of electric field+; and of free charges on the drops,
is now being atudied by the author. Allan been found that although
droplet of '.0-1001.. diameter may rebound after striking an uncontamin-
ated plane eurface of water or a much larger drop, they can always be
made to coalesce by applying vertical fields of only about 10 volts
per cm. The droplets become distorted am they approach the water
eurface, small protuberances develop at their ends in the direction
of the field, and complete coalescence occurs, probably because small
electric discharges occur at the protuberances and cause rupture of
pproved for Release: 2017/09/11 C06461858imer
Approved for Release: 2017/09/11 C06461858
-20-
the intervening air film. This suggests, lerhaps, that once fields
of about 10 V/cm are established in the cloud, coalescence between
colliding cloud droplets and raindrops is assured. In this case
the Elster-Geitel clnrging mechanism will cease long before fields
strong enough to initiate lightning are produced.
At the present time we are unable to suggest a mechanism that
would convincingly account for the origin of lightning in clouds
consisting wholly of liquid water. But there is strong evidence
in favour of the riming-hailstone mechanism being the dominant
one in the typical large thunderstorm.
'proved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
REFERENCES
ATLAS, D., and LUDLAM, F.H. 1961. Quart. J. Roy. Met. Soc.,
88, 117.
BROOK, M. 1958. Recent Advances in Thunderstorm Electricity, p 383.
New York; Pergamon Frees.
BRO.NING, K.R., and MASON, B.J. 1963. Quart. J. Roy. Met. Soc., (in
press).
DUBBIN, W.G. 1956. Air Ministry, Met. Res. Cttee., M.R.P., Ne991.
EIGEN, M., and DE MAEYER, L. 1958. Proc. Roy. Soc., /11.41, 505.
ELJTER, J., and GEITEL, H. 1915.
Phys. Z., 14, 12g7.
EVAI,S, D.C. 1962. M.Sc. Thesis, Durham University.
FINDEISEN, W. 1940. Met. Z. 2z, 201.
1943. ibid 60, 145.
FITZGERALD, D. and BYERS, H.R. 1962. University of Chicago
Contract Report AFCRL/TR/62/805.
GILL, E.W.B. and ALFREY, G.F. 1952. Nature, Lond., 162, 203.
HUTCHINGON, W.C.A. 1960. Quart. J. Roy. Met. Soc. 86, 406.
KRMER, C. 1948. Neth. Met. mat. Vehr. A54, 1.
KUr.TTNER, J. 1950. J. Met., Z, 522.
LA:NAM, J., and MA.011, B.J. 1961a Proc. Roy. Soc., A260, 523.
1961b 1Lid 537.
1962 ATTIT, 587.
LUEDER, H. 1951a Z. angemr: Phys. 247.
1951b ibid 2, 288.
MALAN, D.J. and SChONLAND, B.F.J. 1951a Proc. Roy. Soc. A206, 145.
1951b ibid A2.22, 158.
MASON, n.J. 1953 Tellus, 2, 446.
1957 The Physics of Clouds. (Clarendon Press, Oxford)
p 402-4.
MASON, B.J., and MAY 'ANK, J. 1960. quart. J. Roy. Met. Soc., 86, 176.
r.EINHOLD, H. 1951. Geofis. put% appl. 12, 176.
REYNOLDS, S.E. 1954.
Inst. Min. Tech.
REYNOLDS, S.E., 2G0K, m., and GOURLEY, M.F. 1957. J. Met. 14, 426.
SAnOR, J.D. 1961. J. Geophys. Res., 66, 831.
WALTON, W.H. and IRE.4ETT, W.C.. 1949. Proc. Phys. Soc. B62, 341.
wEICKMANN, H.K. and AUFM KAMFE, H.J. 1950 J. Met. 2, 404
1953 Ibid 10, 204.
Comp. of Thunderstorm Elect.
p 77. New ilex:
proved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
water
supply
spinumg-top
atomiser
cold room
vertical �
wind tunnel
sUnulitted
hailstone
airstn.ten=
vibrating
reed
electrometer
FIGURE I Apparatus for measuring the charge on hail pellets growing by riming.
40 80
drop diameter (p)
HE 2 The !induct ion of tee splinters ( x ) and electric charge (0) as a function of the
diameter of the freezing droplets. Air temperature, � 15 '(:; air voloeity 10 tn/a.
100
miminApproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
�
charge per drop (10-4 e.s.u.)
number of splinters per drop
air-stream velocity (m/a)
notran 3 Splinter ( x ) anti charge (0) production as a function of the impact velocity of
the droplets. Air temperature, �15 �C; drop diameter, 70in
air temperature (�C)
FIGURE 4 Splinter ( x ) and charge (0) production aa a function of the air temperature.
Air velocity 10 in/s; drop diameter, 70a.
pproved for Release: 2017/09/11 C06461858
SESSION 7.1
Approved for Release: 2017/09/11 C06461858
1.
THE THEORY OF LIGHTNING
D.J. Malan
Bernard Price Institute of Geophysical Research
Johannesburg.
The title of this discussion covers a very wide
field since a discussion on any aspect whatsoever of the
lightning discharge or its manifestations involves a certain
amount of theorizing.
I shall confine this talk firstly to the theory
Of the stepped leader, and shall then discuss discharge
processes taking place inside and above the cloud.
Experimental data on the latter processes are
extremely scanty so that the suggested mechanisms are
mostly speculative and as such fall in the realm of theory.
THE THEORY OF THE STEPPED LEADER.
Very little new information based on photography
of the stepped leader or a lightning flash has materialized
since its discovery about thirty years ago. Photographic
studies of laboratory sparks on the other hand have yielded
a large amount of new information but difficulties arise
when attempts are made to apply the knowledge gained to
the leader of a lightning stroke. A case in point is the
recent work of Stekolnikov and Shkilyoli (1962) who used an
Image converter tube to photograph negative rod-to-plane
sparks. They state that the bright steps they photographed
are/
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
2.
are impulse corona discharges and are not preceded by a
pilot streamer, from which they conclude that the leader
process in laboratory sparks is different from that of
natural stepped leaders.
Apart from the observation that the bright step
is not accompanied by a pronounced steplike electrostatic
field change, electrical studies of lightning have given
little information which can be usefully applied to stepped
leader theory.
In studying the radiation fields of flashes in
the distance range of 50 km., both Clarence and Malan
(1957) and Kitagawa (1957) found that when the stepped
leader approaches the ground, radiation pulses follow one
another at intervals of the order of 10/Asec. Kitagawa
ascribes these pulses to short interval loader steps.
Clarence and Malan, however, conclude that since the radiation
field comes from the whole lightning channel, intracloud
streamers which take part in supplying current to the
advancing leader also contribute to the radiation pulses.
The evidence for this conclusion is based on the observation
that of the numerous stepped leaders photographed by us,
none except perhaps the last step, show intervals as short
as 10/Asee between steps. Furthermore, we also argued that
because exactly similar short interval pulses are often
observed to precede strokes subsequent to the first, it is
evident that all the pulses need not necessarily originate
in a stepped leader process.
The question now arises: which explanation of the
profuse /
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 006461858
3.
profuse radiation pulses is the correct one? The answer
to this question is important from the point of view of
stepped leader theory. It is especially necessary to have
conclusive evidence showing whether dart leaders are prop-
agated by short steps or continuously.
The latest attempt to formulate a stepped leader
theory known to the speaker is that of Wagner and Hileman.
The theory of these authors is a sort of combination of the
well-known earlier theories of Schonland, Bruce and
Komelkov. Wagner and Hileman distinguish between the
channel which consists of a 2 mm. diameter highly conducting
arc plasma and a surrounding corona sheath of diameter about
30 m. When the channel is arrested a space charge develops
in front of it being fed by a multitude of filamentary
streamers. At some instant the current in one of the
filaments reaches a critical value of 1 ampere when an arc
plasma begins to develop from the tip of the channel. This
filament now robs the other space charge filaments of their
charge and omergoe as the new channel (or step).
According to Wagner and lilloman the intensity of
the current at any point along the advancing step rises
suddenly when the tip of the newly formed arc plasma reaches
that point and thereafter remains constant until the tip
has reached its maximum extension. The current which
remains at about 1 ampere at the starting point of the step
increases uniformly along the new channel to reach a maximum
value of about 7000 amperes at the tip of the fully
developed /
Approved for Release: 2017/09/11 006461858
Approved for Release: 2017/09/11 C06461858
4.
developed step. At any point the product of current and
duration of current flow is constant during the advance of
the step.
The assumption that the current which flows in
the plasma channel during step formation is mainly derived
from the surrounding space charge can explain why the
bright step produces practically no sudden electrostatic
field change although it carries a heavy average current.
THE POSSIBLE EFFECT OF NEATI RAIN ON THE STEPPED
LEADER PROCESS,
It has been postulated that the preliminary
electrical breakdown in the base of the cloud takes place
by the mechanism of filament formation on large water drops.
In a heavy thundershower the base of the cloud virtually
extends down to ground level. The question now arises as
to what extent, if any, heavy rain is likely to affect the
stepped leader breakdown to earth.
Let us first consider the information available
from the radiation fields occurring immediately before
first strokes of flashes to ground. When using very high
amplification, radiation pulses which may be ascribed to
stepped leaders) can be detected in practically all cases.
For a very largo percentage of flashes the radiation pulses
a
are
surprisingly small, however. Does this indicate that
the breakdown process for flashes in rain is a hybrid
between the ordinary stepped process and the filamentary
process, or can the effect be wholly attributed to the low
radiation /
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
5.
radiation propensities of stepped leaders in general, or
as first suggested by Pierce, are stepped leaders totally
absent?
Let us now consider the photographic evidence.
Under favourable conditions for photography all flashes
show stepped leaders. By favourable conditions it is meant
that the flash Is not obscured by
light scattered from rain has not
of the picture. It is obviously
photograph a faint stepped leader
rain or that the stray
caused overall blackening
almost impossible to
in heavy rain, except
perhaps when the flash is very near indeed.
Berger has taken many pictures of flashes in close
proximity but has only obtained pictures of stepped leaders
on very few occasions. It will be interesting to hear
from professor Berger whether the absence of stepped leaders
on so many of his photographs bears any relation to the
intensity of rainfall while taking the photographs.
THE MECHANISM OF INTRA-CLOUD DISCHARGES.
It can be confidently concluded that intra-cloud
discharges are not propagated by a stepped leader process
because these discharges do not emit the radiation pulses
which are characteristic of stepped leaders.
In regions near the base of the cloud where there
are large water drops, the discharge can be propagated by
the process of filament formation when the field reaches
10 kV/cm, provided that the diameters of the drops exceed
2 mm. At high altitudes in the cloud where its content
may/
t,
_Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
6,
may consist of ice particles and droplets smaller than 2 mm,,
or of ice particles alone, this process can no longer take
part in propagating the discharge. The following is a
tentative explanation of the mechanism of the discharge
process at high levels.
The droplets and also ice particles, as Chalmers
(1947) has shown, will become polarized in the electric
field. When the field between a neighbouring pair of
droplets or ice particles reaches a value of about 30 kV/cm
(the exact value depending on the pressure), electrical
breakdown occurs. The discharge is subsequently carried
forward from particle to particle, probabLy in a channel
of large cross-section.
The presence of discrete pockets of high charge
density in fairly close proximity to each other will cause
the tip fiold to increase rapidly with the advance of the
streamer so that it Is not arrested as in the case of the
stepped leader to ground. Furthermore, if this is the case
it may not be necessary for the tip field to build up to
60 kV/cm which, according to Schonland, is the field required
to produce the thermal ionization required for the progress
of a stepped loader,
PROCESSES CONTRIBUTING TO INTERSTROKE FIELD CHANGES
re,
There are Several factors which can contribute to
the change of electric field as observed at ground level
during the intervals between the strokes of a flash to
ground.
Effect /
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
t 7.
Effect of space charge below the cloud.
A re-arrangement of apace charge between the base
of the cloud and the ground will influence the field in
close proximity to the flash. However, since the conduct-
ivity of the air near ground level is small the relaxation
time is long which makes it unlikely that this effect will
contribute noticeably to the field changes observed during
the short time intervals between the successive strokes of
a flash.
Effect of transient changes in the charging process.
The separation of charge in the cloud under the
influence of gravity is also relatively slow. It is bound
to be profoundly affected during the development of the
discharge, but to what extent is difficult to judge owing
to the lack of experimental data. Moore, Vonnegut at al.
(1t.(2) found by Radar observations that the large drops
responsible for gushes or rain following after lightning flashes
wore not present prior to the electrical discharge. They
suggest that after a return stroke the droplets in the
streamer channels and the surrounding droplets are oppositely
charged so that coalescence takes place by electrostatic
attraction. This observation will have to be taken into
account when formulating a theory relating to the transient
effect of a lightning flash on the normal separation of
charge in a cloud.
The J process.
The interstroke J process is considered to be a
positive streamer discharge progressing mainly upwards from
the/
1
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
8.
the upper regions of the previous return stroke channel.
It was found that at distances nearer than 5 km interstroke
field changes were nearly always negative in sign, whereas
at distances between 12 and 20 km there were very few
negative field changes, about 2/3 of them being positive
and 1/3 zero.
In the intermediate range between 5 and 12 kiss
distance the tendency was for the initial interstroke field
changes of a flash to be positive changing to negative
between the later strokes.
The above distribution of sign of interstroke field
changes at short range are in support of the suggested
mechanism.of the J process.
Effect of continuous discharge to ground
A continuous discharge to ground during the intervals
between the intermittent strokes of a flash brings negative
charge to earth and will thus contribute a positive component
to the interstroke field change at all distances.
Brook, Kitagawa and Workman (1962) have carried
out a detailed study of continuing currents by simultaneous
electrical and photographic observations in New Mexico.
They find that a continuing current flows to ground in
about 25% of interstroke intervals.
THE INTERSTRONE FIELD CHANGES DUE TO DISTANT FLASHES.
At distances beyond 20 km, J field changes although
small should be positive so that it would have been expected
that most of the interstroke field changes in this range
would /
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
9.
would be positive. Such is not the case, however.
In England,Pierce (1955) found that only 25% of
such interstroke intervals showed positive field changes
whilst in the remainder no field changes could be detected.
The above percentage of positive field changes agrees with
the frequency of occurrence of continuing currents in New
Mexico, so that it may be aesumed that large positive
interstroke field changes observed at groat distances are
mainly caused by continuing currents in the channel to
ground. The contribution by the 3 process may be so small
as to be undetectable.
At Johannesburg the distrilpution of sign is somewhat
different. At distances between 25 and 100 km, 19% of
interstroke field changes are positive. Here too, these
field changes may be ascribed to continuing currents. Of
the remaining field changes, however, 37% are zero and 44%
negative.
Pig. L shows three examples of field
changes of flashes in the 50 km range. In (11) a positive
intoratroke field change is followed by 4 negative interatroks
field changes. All the field changes are relatively small
compared with those due to the preceding return strokes.
The sequence of sign in this type of flash is not random
but follows 1,1 the order positive, zero, negative, except
for the occasional random occurrence of an exceptionally
large positive field change which is obviously due to
continuing current.
Of far lesn common occurrence are the field changes
shown in (6) nnd (e) where negative interstroke field changes
are comparable in amplitude with, and may even surpass, the
preceding return stroke field changes. The occurrence at
large /
tpproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
10.
large distances of negative interstroke field changes at
Johannesburg and not in England is possibly connected with
the large vertical extent of thunderclouds at the former
locality which is situated on an extensive plateau 1800
metros above MSL.
As outlined above, the basic data on which to base
a theory for explaining the negative field changes are
scanty but IL appears thnt the effect can be explained in
terms of a readjustment of space charge above the top of
the cloud. A possible theory will now be briefly outlined.
THEORETICAL STUDY OF SPACE CHARGE EFFECTS
Consider unit volume of air above a thundercloud
at an altitude
above the ground.
Let the conduction current be t
; , the
potential cp , field E , conductivity a- and
donsity of charge
since
and
�
if the field is not too high
V
it follows that
giving
99
= crE
- -71" f
- grad 92
- div
(1)
471. f= (i/cr-)oNv j + (3-rad Vcr). (
for steady state conditions, div iv 0
so that
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
11.
4--Tr f = 'AT). (r E)
since the conductivity is constant in the horizontal plane,
we get
f = (i/coi
(2)
Gish (1944) has shown experimentally that in fair weather
'/
can be empirically expressed as the sum of three
exponential terms. Two of these can be neglected at high
altitudes, so that for the region above the cloud his
equation may be written
-o42.3
Vr 0.37 e X lair
ohm. cm, 3, being in km.
If it is assumed that the thundercloud does not alter the
conductivity of the air above it (Gish and Walt 1950),
equation (2) becomes
0.12 E
(3)
A similar expression was derived in a rather more lengthy
way by Holzer and Saxon (1952)
It follows then that the space charge density at
a point In the air above the cloud is proportional to the
electric field at that point.
Using equation (3) it is possible to obtain an
approximate estimate of the distribution of space charge in
a
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
12,
a vertical colunn above the cloud and to calculate the field
change producod at a point on the ground by the relaxation
of space charge after the occurrence of a stroke to ground,
Tamura (1954) expanded the theory of Holzer and
Saxon to prove that the difference in the recovery curves
after an intra-cloud discharge for near and distant flashes
could be explained by taking into account the readjustment
of space charge above the cloud after the occurrence of a
flash, Tamurals expressions for the field component of
the space charge will also be used in what follows.
The interstroke field change duo to the space
charge effect will be called the U field change. A
stroke to ground removes negative charge from the cloud so
that E, of equation (3) becomes more positive and the
a
U field change will consequently be negative at all distances.
To circumvent the uncertainty in estimating the
magnitudes of the charges involved in the discharge processes,
the ratio U/J or the respnctIve field changes wau calculeted
assuming arbitrary values for the charges. Purthermore,
It was assumed that the ratio U/J is unity at a distance of
20 km, Justification for this assumption depends on the
experimental observation that at distances up to about
20 km the interstroke field changes correspond in sign with
the expected J field changes whereas at larger distances
this in no longer the case.
The variations of the ratio U/J with distance,
starting from 20 km, are shown in figs, 2 and 3.
In fig. 2 the J process advances from 4 to 5 km,
altitude and in fig, 3 from 7 to 8 km, The former thus
represents conditions at the commencement and the latter
those at the end of a flash with several strokes.
Curves /
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
13�
Curves A have been calculated on the simple
assumption that the U charge disappears from the top of the
cloud at 12 km altitude as first tentatively suggosted by
Malan and Schonland (1951). Since these curves are
approximately parallel to the abscissae they show that
removal of charge from such a low altitude does not account
for negative interstroke field changes.
Curves B were calculated by the simplified method
outlined above and represent the effect of the adjustment
of space charge in a column reaching from 20 to 30 km
altitude and taking place after a stroke.
Curve C was calculated from Tamurals equations
which assume that th,3 space charge is distributed from the
top of the cloud to infinite height.
The difference between curves B and C is not
significant except in the range 20 to 30 km where curve B
seems to fit the experimental observations somewhat better.
Roth curves show that at large distances the neg-
ative U field change can be much larger than the positive
.7 field change.
E, of equation (3) increases in a steplike
fashion after each partial discharge with the result that
the U field change also increases. This can account for
the observation that the interstroke field change is often
positive or zero after the initial strokes and becomes
negative after the later stroke or a flash.
As Brook, Eitagawa and Workman have pointed out
the 3 field changes become very small at a distance of 50 km
from /....
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
14,
from which it follows that the curves of figs. 2 and 3
do not account quantitatively for the relatively large
negative field changes shown in fig. 1 (i) and (C.
It is possible that the field above the cloud
occasionally becomes so large that equation (1) is not
valid and the outlined theory breaks down. This may
conceivably happen In a cloud of large vortical extent
where the main positive and negative centres are widely
separated so that flashes to ground are more frequent than
intra-cloud flashes. When the records shown in fig.l
B and C were obtained flashes to ground were 2 to 3 times more
frequent than cloud flashes.
In the extreme case, the field above the cloud
may benome high enough to initiate a glow discharge between
the space charge end the ionosphere (Malan 1037). In this
ease the conductivity of the air above the cloud will increase
and cause a decrease in the relaxation ttmo thus giving
large and rapid U field changes.
It will be interesting to determine whether
the rare large negative field changes have any connection
with solar flares whichcause increased ionization in the
D layer of the ionosphere.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Ii
4
REFEHENCES.
Brook M., Kitagawa N. and Workman E.J. 1962
Quantitative study of strokes and continuing
currents in lightning discharges to ground.
Journ. Geophys. Res. 67, 649
Chalmers J.A. 1947
The capture of ions by ice particles.
Q. Journ. Roy. Met. Soc. 73, 324.
Clarence N.D. and Malan D.J. 1957
Preliminary discharge processes in lightning
flashes to ground.
Q. Journ. Roy. Met. Soc.. 83, 161.
Gish 0.11. 1944
Evaluation and interprotatior. of the columner
resistance of the atmosphere.
Torr. Magn. and Atm. Eloctr., 49, 159,
Gish 0.11. and Wait G.R. 1950
Thunderstorms and the earth's general electrification.
Journ. Geophys. Res, 55, 473,
Holzer R.E. and 2axon D.. 1952.
Distribution of t!lectrical conduction currents in
the vicinity of thunderclouds.
Journ. Geophys. Res. 57, 207.
Kitagawa
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Ritagawa N. 1957
On the electric field change due to the leader
processes and some of their discharge mechanisms.
Papers in Meteorology and Geophysics vol.VII, no.4
published by the Met. Res. Inst., Tokyo.
Kitagawa N., Brook M. and Workman E.J. 1962.
Continuing currents in cloud to ground lightning
discharges.
Journ. Geophys. Res. ,62, 637.
Malan D.J. 1937
Sur los d6charges orageuses dans la haute atmosphere
Comptes Rendus 205, 812.
Malan D.J. and Schonland B.F.J. 1951.
The electrical processes in the intervals between
the strokes of a lightning discharge.
Proc. Roy. Soc,A, 206, 145.
pore C.B., Vonnogut B., Machado J.A. and Survilas H.J. 1962.
Radar observations of rain gushes following overhead
lightning strokes.
Journ. Geophys. Res. 67, 207.
Pierce E.T. 1955.
Electrostatic field changes due to lightning discharges
and The development of lightning discharges.
Q. Journ. Roy. Meteoro1. SOC. 81, PP. 211 and 229.
Stekolnikov /
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Stekolnikov I.S. and Shkilyov� A.V. 1962.
New data on negative spark development and
its comparison with lightning.
Conference on Gas Discharges and the
Electrical Supply Industry.
In course of publication by Butterworth.
Tamura Y. 1954.
An analysis of electric field after lightning
discharges.
Journ. of Geomag. and Geoelectr. 6, 34.
Wagner C.F. and lineman A.R.
The lightning stroke II
(Typed memorandum. Has it already been
published ? )
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
_..---------'
A
B
C
I
I 1
a 100
millisec.
I
200 300
FIG.1.
i pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
cn
1/n.-
40
a
kilometres
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
SaJlaWO]pi
I\)
0
01
03
0
0
en&
ru
C)
CD
0
Approved for Release: 2017/09/11 C06461858_____
Approved for Release: 2017/09/11 C06461858
.SASION 7.2
Types of Lightning
Introduction
N. Kitagawa
The mechanism of lightning discharge can best be studied by the
simultaneous measurement using photographic and electric-field record-
ing technique.
Lately in New Mexico lightning measurements on this
line with improved technique have been done extensively. The result
thus obtained have revealed new aapects of lightning discharges as
well as the detailed structure of the discharge mechanism. (Kitagawa,
Brook and Workman, 1962; Brook, Kitagawa and Workman, 1962)
Based mainly upon these results, the author will try to depict
the picture of the lightning, pointing the accompanying unsolved
problems at the same time.
Terminotsgy
The author adopts the terminology of Siionland (1956) in describing
! the various processes in the lightning discharge. A flash is a light-
1
1 ning discharge in its totality. A stroke is a partial dischange con-
f
I
4:
sisting of downward-moving return streamer. A flash may consist of
a single stroke or a series of strokes in the same or an adjacent
channel. A M component is a sudden enhancement of the continuing
luminosity which occasionally follows a stroke in the channel (Malan
and 4honland, 1947). The M components are not preceded by leaders.
Long-continuing luminosity is arbitrarily defined as luminosity
which persists in the channel for a time longer than 40 .sec i.e.,
luminosity which lasts as long as or longer than the usual stroke
1
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
)
�
c�-�:1�44:
interval. A stroke followed by such luminosity will be called a long-
continuing stroke. Occasionally a stroke is followed by continuing
luminosity which lasts less than 40 msec; such a stroke will be called
a short continuing stroke. A stroke whose luminosity decays abruptly
will be called a discrete stroke. A K change is a small, rapid elec-
tric field change which occurs in the intervals between and after the
strokes of a multiple stroke flash ( Kitagawa, 1957; Kitagawa and Brook,
1960 ). The K changes are generally associated with streamer activity
within the cloud. There is a slow electric field change which occurs
during the interval of continuing luminosity, this will be referred to
as a continuing or C change to distinguish it from the junction or J
change which occurs without the accompanying channel luminosity between
or after the strokes.
In analogy with the difinitions given by Malan (1954), a lighting
flash which involves one or more continuing strokes is called a hybrid
flash. A flash which involves discrete strokes or short continuing
strokes is called a discrete flash. Cloud-to-Ground, in-cloud and
cloud-to-cloud discharges will be referred to by the symbols, C-G,
I-C and C-C discharges respectively. A cloud-to-clearair flash is
called a air discharge.
General nature of C-G, discharges
AS a result of New Mexico lightning measurements it has been found
that hybrid flashes i.e. flashes involving one or more long continuing
strokes are found to be observed very commonly in G-4 discharges.
- 2 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
In multiple flashes which constitute 86 per cent of all C-G discharges,
the occurrence rate of discrete and hybrid flashes is about fifty to
fifty. The percentage of single flashes with long continuing luminos-
ity is exceptionally low (only 2 of 193). The long continuing strokes
does not either occur as the initial stroke of a multiple flash.
Figure 1 is a comparison of the luminous events of a discrete and a
hybrid flash as recorded by the moving-film camera and by the electric
field and electric field-change records. In the schematic represen-
tation of the luminous events, a straight, vertically oriented channel
is assumed. The electric field record represents the actual varia-
tion of electric field, whereas the electric field-change record
emphasizes the rapid components through the use of an antenna with a
short time constant and a high amplification.
Both of the flashes
are about 20 km dimtant from the recording station. It can be noted
in Figure 1 that the magnitude of slow electric field changes ( changes)
is very small or practically zero during the intervals of non-luminosity
between strokes of both discrete and hybrid flashes. On the contrary,
it has been found that a large positive, slow electric field change
c
change) is always associated with a continuing luminosity on the
photographic record.
The average munber or strokes per flash in
both discrete and hybrid flashes is 7. If we include the single
flashes, the average number of stroke per flash is 6. Thus the num-
to
ber of strokes per flash has foundAbe appreciably larger than the sta-
tistics by Schonland (1956) based upon electric field-change records.
It is not uncommon to find several strokes which produce field-changes
- 3 -
a pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
of 1/10 to 1/20 of the field change caused by the largest stroke.
Without the positive identification of the leader-return combinations
on the high speed photographs, it is highly probable that R changes of
such small magnitude might have been interpreted as K changes or over-
looked.
Malan (1954) found that the large slow field change of the same
nature as defined C change here, often occurs as a final stage of a
multiple flash which involves fewer stroke elements. In New Mexico
measurements such a hybrid flash found to be observed more frequently
that its earlier stage is very simOillar to that described by Malan
but involves one or more stroke elements of very small R changes in
its very late stage.
In England Pierce (1955 a) sometimes recorded s(p) field change.
The field change of this type can reasonably interpreted as a conti-
nuing current field change in which the stroke field change initiating
the C change is barely discernible on the electric field record.
It is because of the occurrence of such small stroke field changes that
the number of strokes per flash appears to be less when counted on the
electric field records than when measured on the high speed photographs.
While the duration of long continuing luminosity varies widely
as shown in Figure 3, the duration of the no-luminous interval i.e.,
o-f ptvrf dt-vift �.fruke
the interval from the end of the luminosityAto the following stroke
tends to fall in certain limited range around 80 msec (from 50 to 200
msec). Though the duration of a no-luminous interval appears to be
longer than a usual stroke interval, the value still lies within the
- 4 -
a pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
range of discrete or short stroke intervals (5 to 180 msec).
The methods and assumptions used in calculating value for the
charge which is brought to earth by individual discharge elements of
a flash are described and discussed in detail by Brook, Kitagawa and
Workman (1962).
Here the method will be outlined. The electric
field change due to a lightning stroke measured at the surface of
the earth is given by
where
E
QR
_
D.! 1.4 p, ))/
AE is in volts/cm, QR is in coulombs and HR and D are in km.
D is the horizontal distance from the field meter to the flash and
HR is the vertical height to the assumed center of chaoge QR .
The bight HR is determined by
R =.=
h
Where h is cloud base hight, te is the total duration of the dart
leader measured on electric field change records and tp is the time
for the dart leader to travel from the cloud base to ground determined
by photographic records. From the first equation, with the measured
value of IIR , D and AE, the charge QR can be written
OR -
E CD2t th te /t -.13/2
53lO h te t
The above emthod is also applicable to the continuing current intervals.
Let Hi and Hz be the heights determined for two successive strokes
between which a continuing current to ground was evidenced on both the
electric field records and the photographs. The charge Qc is assumed
- 5-.
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 006461858
to be centered at HU , a distance midway between the tops of the two
return stroke channels; i.e.,
= HI + Hi)
The corresponding slow electric field change 4 E is then measured,
and the charge Qc is calculated.
The negative charge (no strokes carrying positive charge to earth
were observed) lowered by individual stroke is shown in Figure 2(a)
and 2(b) in two histograms showing the number of occurrences of strokes
in which charge was (a) lowered by strokes which were preceded by
stepped leaders and (b) lowered by strokes preceded by dart leaders.
A striking contrast is seen to exist between the two types of strokes.
The minimum charge lowered by the strokes associated with stepped
leaders is 3.0 coul; for the others a minimum value is 0.21 coul.
Both of the histograms exhibit a sharp cut-off at their minimum end.
The most frequent value of charge brought down by first strokes lies
between 3 and 4 coul; for subsequent strokes the most frequent value
lies between 0.5 and 1 coul. For single flahses the average value
of the charge lowered was calculated to be 4.6 coul. The charge
lowered by the continuing current is shown in Figure 3 as a plot vs.
its duration.
The maximum duration of a continuing currnett interval
was found to be 300 msec. The greatest amount of charge lowered
during one of these long intervals was 31.2 coul. Though both the
amount of charge and the duration vary widely, the plot shows that
the average current i.e.,the charge devided by the duration tends to
be far leas variable from current to current. It ranges from 38.4
- 6 -
pproved for Release: 2017/09/11 006461858
Approved for Release: 2017/09/11 C06461858
to 130 amp around the mean value of 79.2 amp. Thus the continuing
current turns out to be very efficient agent for carrying the cloud
charge to earth. Because of this agent the total charge lowered by
a hybrid flash is remarkably larger than that lowered by a discrete
flash. The calculation shows that the average values for a discrete
and a hybrid flashes are 20 and 34 coul respectively.
The result of the calculation of the charge center hight for each
stroke element is shown in Figure 4 as a plot vs. the stroke order.
The figure shows the definite tendency that the hight of charge center,
exactly speaking, the length of the stroke channel increases from
stroke to stroke. The most frequent height difference for the dis-
crete flash is 0.3 km. The value for the discrete intervals of the
hybrid flash is also 0.3 km. The continuing-current intervals are
most frequently associated with a value of 0.9 to 1.6 km. Figure 4
suggests that hybrid flashes usually involve greater cloud volumes
than do discrete flashes.
Continuing currentsand the junction Process
z
With the realization that the continuing currents to earth often
occupy the intervals between strokes previously assigned to in-cloud
processes alone (i.e. J changes), it is desirable that we reexamine
the interpretation of the inter-stroke field changes, as discussed by
Malan and Schonland (1951 b), Malan (1955) and Pierce (1955 a,b)
Table 1 shows the electric moment change 211Q associated with a
flash, a stroke, a long continuing current and a J process, obtained
- 7 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
in the present measurement. J change, identified by the absence of
Table 1 Electric Moment Change Associated with a Flash,
a Long Continuing Current and a J process
Discharge Electric Moment Change 2HQ (coul km)
Flash 248 (average)
Stroke 22 (average)
Long Continuing Current 135 (average)
Junction Process 1.62 (maximum observed value)
a continuing channel luminosity on the photographs turns out to be
remarkably small. Asshown in Table 1 the maximum change was found
to be 1.62 coul km. Large slow field changes do occur, but
invariably the photographs show these to be associated with continu-
ing current to ground. When we consider the average change in
moment associated with the flashes in this study is 248 coul km
(Table 1), it is not surprising to find that the J-change moment,
having values much less than 1.6 coul km, are not detectable.
Taking the maximum value as 2 coul km, we see that the J process
produces a change in moment which is about 10 per cent of the average
change in moment for strokes (22 coul km), and about 1 per cent of the
- 8 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
average change in moment for continuing currents (135 coul km). It
is now clear that the value 2HQ = 50 coul km used by Pierce (1951 b)
for the J change in moment is really to be associated with the C change.
We shall assume that the most favorable conditions of noise allow
the measurement of slow field changes of magnitude 10-3 of the earth's
fair weather field, i.e., of magnitude 10-3 volt/cm. Using the value
1.6 coul km for the upper limit of the J change in moment, we calculate
that the maximum distance at which a J change will produce an electric
field change of 10-3 volt/cm is 52 km. Since only a C field change
is expected to be detected as a slow electric field change on the rec-
ord beyond this distance, it is highly probable that most of the J
changes reported by Pierce (1955 a) for distances between 50 and 90
km were produced by continuing currents to ground. He was able to
measure slow field change in approximately 25 per cent of the inter-
*FY
vale between
be detected.
with our own
strokes and for the remainder no
This figure of 25 per cent is
statistics for the occurrence of
variation in field could
reasonably consistent
continuing currents in
lightning discharges. These statistics also reinforce the conclu-
sions that C changes, and not J changes, are detectable for distances
beyond 50 km. A new process involving a discharge from the cloud
top to the high conducting layers was postulated by Malan and Schonland
(1951 b) and Malan (1955) to explain the apparent absence of J changes
in the measurements of Pierce (1955 a) and Malan (1955) for distant
storms (20 to 150 km). Since we now see that the absence of J
change is actually to be expected for distances beyond 50 km, the com-
- 9 -
1;
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
pensating process (a discharge to the upper air) is unnecessary.
Also, it appears that it is not nesessary to postulate a difference
between thunderclouds in England and in South Africa (Pierce and
Wormell, 1953).
During the interval between successive strokes of a multiple
flash, there found to be two different stages of channel conditions,
a continuing luminosity stage and a non-luminosity stage. We have
done some experimental approach to the luminous conditions, but the
channel condition during no-luminous intervals remains entirely un-
known. A number of problems concerning this channel condition
should be the subjectsof future studies. For instances, what is
the amount of the dark current in the channel? How the J process
is connected to the ground? How such a conductive condition is
maintained in the channel that allows the dart leader of a subsequent
stroke to follow the same channel.
M components and K changes
During the continuing luminosity M components are found to be
associated with field changes similar to K changes on electric field-
change records as can be seen in Figure 1. Seperations of M compo-
nents are very small and tend to increase very rapidly with elapsed
time within the first 15 msec from the return stroke.
Later on this tendency desappears and M-component intervals exhibit
no dependence on the elapsed time.
Figure 5 (a) shows the frequency
histogram of M-component intervals in this later stage of continuing
- 10 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
41.
luminosity. For the comparison the frequency histograms for K changes
in discrete intervals of G-C discharges (b) and for K changes in I-C
discharges (c) are shown in the figure.
So far the luminosity continues in the channel, streamers connected
to the channel keep developing within the cloud into the flash charged
region. The M components is the presentation of the current surge
in the luminous channel produced by the momentary increase of the
cloud charge supply. Thus, the occasional appearance of M components
in a continuing current is considered to be the reflection of the non-
uniform distribution of charge in the cloud. The similarity in
three histograms (a),(b) and (c) in Figure 5 also suggests that M-
component intervals and K-change intervals both in discrete stroke
intervals of C-G- discharge and in later stages of 1-c discharges are
all controlled by the same conditions, by the conditions attributable
to the cloud structure, not by the conditions of the discharge process.
Ogawa (1962) calculated the average developing velocity for the stream-
ers associated with the long continuing current to be 1.6 x lOC cm/sec.
This value, combined with the most frequent M-component interval of
6 msec, glves the linear dimension or 100 m for the spacing of the
densely charged regions in the cloud. This dimension can be com-
pared reasonably well with that of the unit cell of the convection
suggested by Reynolds (1954) i.e., the microstructure in the so-called
cloud cell, evidenced by the radar echoes of the strom or by the visual
observation of cumulus towers and striation in the rain sheet.
1
The K change is a rapid luminous small scale discharge within
r:
tti
&I
pproved for Release: 2017/09/11 C06461858
�.-r.T.^ ^ ,.-�-__
Approved for Release: 2017/09/11 C06461858
the cloud. Usually the duration is leas than 1 msec and the moment
change involved varies from a few hundredth to about 1 coul km.
Ogawa's (1962) analysis based on New Mexico measurements shows that
the main discharge process which constitutes the K change is the rapid
flow of the cloud negative charge into an already existing channel.
Comparing the discharge processes associated with M components with
those produces K changes, there seem to be no essential difference in
the way of the streamer development within the cloud. A junction
or J process described by Malan and Schonland (1951 b) is now reason-
ably interpreted as a whole series of K-change discharges involved
in a stroke interval. The J change turns out to be the smoothed
trace of the electric field record which actually consits of a number
of very small K-change steps during the non-luminosity interval of a
multiple C-G flashes.
Nature of I-C discharge
Figure 6 shows a typical example of electric field and electric
field-change records of a I-C discharges. The records is usually
dpvided into two different portions; the earlier active portion and
the later portion. The later portion of field and field-change
records are very similar to those between strokes of a C-G discharge
i.e., J or C field changes; K changes follow each other on the electric
field-change record with the identical time intervals with those of
K changes and M components in C-G discharges (Figure 5).
During the earlier portion pulse activity is much higher; amplitudes
-.12 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
of some of pulses are much larger and pulses are spaced so densely
that quiescent intervals are seldom recorded. Occasionally an
earlier active portion may be preceded by a less active portion,
called an initial portion, which is characterized by pulsations of
high repetition rate and of relatively small amplitude. The rate
of the moment change estimated on electric field record is generally
higher during active portions than during later portions.
fhe electric field and electric field-change records of a later portion
obviously show that the discharge mechanism is essentially the same
as that of discharge processes which take place in the cloud during
the intervals of a multiple C-G discharge. Though there is no strict
distinction between the discharge of the M-component type and of the
K-change type in the case of a I-C discharge, the discharge mechanism
is considered to be closer to that of the continuing current interval
of a C-G discharge, because the moment change in this portion is gen-
erally larger than that in the discret stroke intervals (i.e. J field
change). And it is highly probable that continuing channels of con-
sidarable length are usually established in a later stage of a cloud
discharge as occasionally evidenced on moving camera records of C-C
or Air discharges. Occasionally a moment variation in a K change is
appreciably larger compared with that in C-G discharges.
A measurement by Ogawa (1902) shows that for the I-C discharge the
average moment change associated with the K change is about 6 coul km
and the charge involved is estimated to be 2 coul.
current during the change is 2000 amp.
- 13 -
The average
�
pproved for Release. 2017/09/1
Approved for Release: 2017/09/11 006461858
A very frequent occurrence of pulses in an active portion appar-
entry suggests that a number of different streamer channels develop
in the cloud at the same time or with slightly different time phases.
Sometimes the development of relatively small scale streamers is
repeated for the duration of some 50 to 100 mdec before the burst of
larger streamers takes place (an initial portion). At present quan-
tative informations about streamer processes in an initial or in an
active period of the I-C discharge are not available. However, we
can point out one aspect of the initial stage of the I-C discharges.
Regardless the discharge starts with an initial portion or immediately
with an active portion, the pulse activity is much more irregular than
that recorded in the very beginning stage of the leader field change of
C-G discharge, both length of pulse intervals and pulse amplitude tend
tobelargerandmorevariableinthefiristseveratamescomwred with
those in the later stage. Still the regularity in both period and
amplitude well reflects the step-wise development of stepped leader
streamers.
In fact the pulse intervals during this stage lies in the
range from 10 to 150 ,sec. On the contrary for corresponding initial
stage of the field change of a 1-C discharge, pulse intervals spread
over the extremely wide range from 10 Alsec,to several msec. Pulse
intervals along with much irregular pulse shapes indicate that the
mechanism of the associated discharge is entirely different. While
the initial breakdown of a C-G discharge takes place in the water
droplets region of a cloud, the initiation of a I-C discharge starts
at much higher altitude where the cloud consists of ice particles and
- 14 -
pproved for Release: 2017/09/11 006461858
Approved for Release: 2017/09/11 C06461858
A very frequent occurrence of pulses in an active portion appar-
entry suggests that a number of different streamer channels deve1o0 1
in the cloud at the same time or with slightly different time phases.
Sometimes the development of relatively small scale streamers is
repeated for the duration of some 50 to 100 maec before the burst of
larger streamers takes place (an initial portion). At present quan-
tative informations about streamer processes in an initial or in an
active period of the I-C discharge are not available. However, we
can point out one aspect of the initial stage of the I-C discharges.
Regardless the discharge starts with an initial portion or immediately
with an active portion, the pulse activity is much more irregular than
that recorded in the very beginning stage of the leader field change of
C-G discharge, both length of pulse intervals and pulse amplitude tend
to be larger and more variable in the first several 01044 compared with
those in the later stage. Still the regularity in both period and
amplitude well reflects the step-wise development of stepped leader
streamers.
In fact the pulse intervals during this stage lies in the
range from 10 to 150 Ak.sec. On the contrary for corresponding initial
stage of the field change of a I-C discharge, pulse intervals spread
over the extremely wide range from 10 Aksec .to several msec. Pulse
intervals along with much irregular pulse shapes indicate that the
mechanism of the associated discharge is entirely different. While
the initial breakdown of a C-u discharge takes place in the water
droplets region of a cloud, the initiation of a I-C discharge starts
at much higher altitude where the cloud consists of ice particles and
- 14 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
super-cooled droplets of very small size. The author suggested that
the difference between the two initial breakdown processes is attrib-
table to the difference in the breakdown impedance affected by the
above two different enviromental conditions i.e., the difference in
the relative populations of water drops and ice particles in the cloud
(Kitagawa and Brook, 1960). In addition to the difference in the
breakdown impedance, it is probable that the kind of breakdown stream-
ers, positive streamers mostly, not negative streamers and the config-
uration of channels, extensively branched or separated into a number
o4 the
of channels will account fot the different character 141----t-h-e--vePy--ear-ky
breakdowin p.atess
stage af-4-11w---444-114--eitradige- of I-C discharges.
We have tried to depict the nature of the I-C discharge in the
comparison with the C-G discharge. As to discharge processes,
however, which take place entirely in the cloud, we have very little
quantative informations. For the further study of the lighting
discharge, quantative measurements of these irccesses are desired, e.g.
r:niitil brehkdo%.n streemer Irocess, h streamer prz.cess in an active
portion. a K-chan.e 1.rocess 4nd a stresmer process LI continuing luminosity.
Lne eri:roech or these .1.1 Le si%ultaneous measurements by field meters
cf high tAse-resolution st several stations on ...he surface.
- 15 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
References
Brook, M., N. Kitagawa, and E. J. Workman, Quantative Study of strokes
and continuing currents in lightning discharges to ground,
J. Geophys. Research, 67, 649-659,1962.
Kitagawa, N., On the mechanism of cloud flash and junction or final
process in flash to ground, Papers Meteorol. and Geophys. Tokyo,
7, 415-424,1957.
Kitagawa, N., and M. Brook, A comparison of intracloud and cloud-to-
ground lightning discharges, J. Geophys. Research, 64, 1189-1201,
1960.
Kitagawa, N., M. Brook, and E. J. Workman, Continuing currents in
cloud-to-ground lightning discharges, J. Geophys. Research, 67,
637-647, 1962.
Malan, D. J., Les decharges orgeuses intermittentes et continues de
la colonne de charge negative, Ann. geophys., 10, 271-281,1954.
Malan, D. J., Les decharges lumineuses dans lea nuages orageux, Ann.
geophys., 11, 427-434, 1955.
Malan, D. J., and B. F. J. Shonland, Progressive lightning, 7,
Directly-correlated photographic and electrical studies of lightning
from near thunderstorms. Proc. Roy. Soc. London A, 191, 485-503,
1947.
Malan, D. J., and B. F. J. Shonland, The distribution of electicity
in thunderclouds, Proc. Roy. Soc. London A, 209, 158-177, 1951a.
Malan, D. J., and B. F. J. Shonland, The electrical processes in the
intervals between the strokes of a lightning discharge, Proc. Roy.
Soc. London A, 209, 158-177, 1951b.
Ogawa, T. Private communication, 1962.
Pierce, E. T., and T. W. Women, Field changes due to lightning
discharges, in Thunderstorm Electricity, edited by H. R. Byers,
University of Chicago Press, 1953.
Pierce, E. T., Electrostatic field changes due to lightning discharges,
Quart. J. Roy. Meteorol. Soc., 81, 211-228, 1955a.
Pierce, E. T., The development of lightning discharges, Quart. J. Roy.
Meteorol. Soc., 81, 229-239, 1955b.
- 16 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Reynolds, S. E., Compendium of Thunderstorm Electicity, (Signal Corp
Project, 172-B), New Mexico Institute of Mining and Technology,
Socorro, New Mexico, 1954.
Schonland, B, F. J., The lightning discharge, Handbuch der Physik,
22, 576-641, 1956.
- 17 -
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
.14
List of Figures
Figure 1
:
Examples of simultaneous photographic, electic field,
and electic field-change records of discrete and hybrid
flashes. Positive deflections are upward.
fr
Figure 2
:
Frequency histograms of (a) charge lowered by strokes
preceded by a stepped leader, and (b) charge lowered by
strokes preceded by a dart leader.
Figure 3
:
Graph of charge vs. duration for the continuing current.
Figure 4
:
Apparent hight vs. stroke order for strokes in discrete
and hybrid flashes.
s.j
Figure 5
:
Frequency histogtams of (a) M-component intervals in
continuing luminosity; (b) K-change intervals in discrete
intervals of C-G discharges; and (c) K-change intervals
4
in later portions of I-C discharges.
Figure 6 : Examples of simultaneous electric field-change (above)
and electric field (below) records of a I-C discharge.
Positive deflections are upward.
- 18-
3
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
�
:*
PHOTOGRAPHIC RECORD
DISCRETE FLASH
R, R,
R,
R. R. R. R,
ELECTRIC FIELD RECORD ����
�
I
.i.. _ ___,114.�.J.--........."\� - i ,-...w._k__ ? \.,.....- ..._^_)�_/,___,-&-__ ___ pf\,- _A._ ,
ELECTRIC FIELD CHANGE RECORD
0 50 100
TIME (msec)
HYBRID FLASH
PHOTOGRAPHIC RECORD
J.,....� 4.� .--..
-
Ft, Rs
I.
CONTINUING CURRENT
ELECTRIC FIELD RECORD
1
R, R. R,
C FIELD CHANGE
R. R, Re
J 1,4 CHANGE
4 \
_ J"....... ..........
ELECTRIC FIELD CHANGE RECORD
0 50 100
TIME (msec)
K CHANGE
pproved for Release: 2017/09/11 C06461858
9991.917900
NUMBER OF OCCURRENCES
0
0 1 2 4 5 6 7 8 9011 12 L3 14 15 16 17 18 19 20
CHARGE LOWERED BY STROKES PRECEDED BY A STEPPED LEADER
(COULOMBS)
0
cJ
1�011.
Alm
411�MINEP
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
CHARGE LOWERED BY STROKES PRECEDED BY A DART LEADER
(COULOMBS)
�cs
�cs
S
)
CD
-h
0
-5
CD
CD
CD
)
CD
cD
Co
9991.917900
. I,
CURRENT
CONTINUING
LOWERED
C.,
Approved for Release: 2017/09/11 C06461858
�
cr
0
0-.-
0
0 50 100 150 200 250 300 msec
DURATION OF CONTINUING CURRENT
pproved for Release: 2017/09/11 C06461858
999 [MOO 1.1./60/LI.OZ :aseaia JOI panaidd
4:2
APPARENT HEIGHT
0 2 4 6 8 10 12
0
40.14
it-gy+.6io+ 6�
N � � V++
,Q -4- 0
� II
+ +
a,-.4.004:14. 0 t-1- 0+
+
0 00 00 00 + + +4- -I-
0 0
0 +
0 0 0 00 + 0 + + + +
173-
-

0 0 Ocp 0+
00 0 0++
S3HSV1A 012:18AH +
cf) +
o o o o + o
00 0 00 ++
0 0 0 4- -4-
0 0 0
0 00 +
00
0
4-
-I- -4-
999 [MOO 1.1./60/LI.OZ :aseaia JOI penaiddV
��� n An- /10.."..�
Approved for Release: 2017/09/11 C06461858
�;$
PERCENTAGE DISTRIBUTION
PERCENTAGE DISTRIBUTION
ae
0
in
L1 , 1
o ,I I I-t- 1-- t----t----
0 2 4 8 12 16 20 24 28 32 msec
ae
0
(NJ
(o> M COMPONENT INTERVALS
�
0 2 4 8 12 16 20 24 28 32 msec
(b) DISCRETE K -CHANGE INTERVALS-- C- G DISCHARGES
0 2 4 8 12 16 20 24 28 32 msec
(c) K - CHANGE INTERVALS-1-C DISCHARGES
pproved for Release: 2017/09/11 C06461858
SPSIoN R.I
Approved for Release: 2017/09/11 C06461858
Lightning Protection
D. Miiller-Hillebrand
Institute of High-tension Research, University of Uppsala,
Uppsala, Sweden
Physical research on lightning and the practical application of its
results are intimately related to each other. The demand for security
against damage by lightning has greatly stimulated research. Lightning
discharges affect buildings, power lines, machines and equipment, and
telecommunications systems. They can cause damage to aircraft and during
tunnel blasting deep down in a mountain. They kill and injure living
creatures and cause fires and accidents. All this has given rise to
innumerable investigations and publications in the majority of civilized
countries in which lightning is a problem. Summaries on lightning pro-
tection matters have often been published. An early work by Goodlet (1)
deals with questions such as the shattering of poor conductors, damage
to buildings, oil-tank fires, damage to aircraft and effects on living
creatures. A more recent work by McEachron (2) gives, in the first ,
place, an account of protective measures for communications and power-
-supply systems. The present report can only give a limited survey of
such protection questions as have still not been completely elucidated;
they relate, on the one hand, to physical phenomena and, on the other,
to statistical and probability investigationstwhich are often connected
with financial problems.
The publications from, the Golden Age of lightning research - from
about 1750 to 1780 - often show acute observation of nature in connec-
tion with protective measures. As an example, I quote two observations
by Benjamin Franklin which touch on quite topical questions. The first
shows his view of the limited protective range of an elevation rod -
a question which is always being discussed. In "Poor Richard's Almanac"
for 1753 Franklin writes:
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
Provide a small Iron Rod (it may be made of the Rod-iron used by
the Nallers) but.of'such a Length, that one End being three or four
Feet in the moist Ground, the other may be six or eight Feet above
the highest Part of the 6uilding. ... If the House or Barn be long,
there may be a Rod and Point at each End, and a middling Wire along the
Ridge from one to the other. ...
The instruction is later repeated in Franklin's 24th letter (3).
The second observation makes clear the protective effect of a thin
wire, in a case which is topical today, as regards finance. The church
at Newsbury in rew England had been badly damaged by lightning in the
summer of 1754. Franklin gives a detailed account of the destruction.
I reproJuce extracts from his letter to Dalibard (3):
The spire (of wood, reaching 70 feet higher than the 70 feet high
stf.eple) was sent all to pieces by the lightning, and the parts flung
in E01 directions over the square in which the church stood, so that
nothing remains above the bell. ... Prom the end of the pendulum, down
ouite to the ground, the building was exceedingly rent and damaged, and
some stones in the foundation wall torn out and thrown to the distance
of twenty or thirty feet.
The central portion was not damaged. The lightning current had
destroyed a thin wire which connected the clapper of the bell to the
clock mechanism 20 feet away. The flash was conducted to the pendulum
wire, which had the thickness of a goose quill. Franklin draws the
following conclusions:
1. That lightning, in its passage through a building, will leave
wood to pess as far as it can in metal, and not enter the wood again
till the conductor of metal ceases. And the same I have observed in
other instances, as to walls of brick or stone.
2. The quantity of lightning that passed through this steeple must
have been very great, by its efVects on the lofty spire above the bell,
and on the square tower all below the end of the clock pendulum.
3. Great as this quantity was, it was conducted by a small wire and
a clock pendulum, without the least damage to the building so far as
they extended.
4. The pendulum rod) being of a sufficient thickness, conducted the
lifThtning without drelage to itself: but the small wire was utterly
destroyed.
5. Theuen the ;mall wire was itself destroyed, yet it had conducted
the liotning with earety to the building.
6. And from the whole it seems probable that, if even such a small
wire had been extended from the spindle of the vane to the earth,
before the storm, no damage would have been done to the steeple by
that stroke of lightning though the wire itself had been destroyed.
Approved for Release: 2017/09/11 C06461858 A
Approved for Release: 2017/09/11 C06461858
'711
- -
riefore concluding this short historical survey, I show (Pig. 1) one
of th arliest liglAning-conductor dhlwitla ever published - a lightning
conductor designed by Torbern Tergman for a building in Uppsala, riweden,
In 1705. Fven at that period the special instruction is given that
laree metal narts in the building should be connected to the lightning
conductor (4)..
The Lightning Path near the Ground
Space charges between cloud and earth greatly affect the lightning
path. A visible proof of this is the "Type p leader" according to
Schonland (5, 6), which has a high velocity (more than 6 x 107 cm/see =
= 600 M/ms) between the base of the cloud and the space-charge layer
(first stage) 3nd a low velocity (about 1 x 107 = 100 m/ms) and often
a ?renounced fork in its further career to the ground (second stage).
About 30W-, of the leaders in South Africa showed this phenomenon (7).
The chnrges in the water-vapour cloud are transported to and distributed
over the seace-charge cloud. These charges consequently become "over-
-neutralized". The leader's space-charge channel receives additional
charge and thereby has a greater volume than it would have had if the
space-charge layer had not existed. On the return stroke the lightning
channel accordingly receives a substantial additional charge. The
current strength is increased. The course of the lightning current,
according to Berger (8), who recorded it on the summit of Monte San
Salvatore, shows that a current maximum is reached after 5-10/4a.
The vertical length of the lightning channel after 5-10)as is approxi-
mately 400-1200 m (5). The order of magnitude of the distance between
the space-charge cloud and th+round agrees with the length of the
leader's "second stage" measured by Schonland et al. At Monte San
Salvatore the space charges may be particularly heavy. The two 70-metre-
-high towers on the 600-metre-high mountain both generate glow dis-
charges in the static electric field. The charges - approximately 1 Coul
in 10 minutes - may affect the lightning path. The statiksics of
1.
piroved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
lightning strokes to ground show this clearly (9). The annual number of
lightning strokes to these towers is 29. if this is converted to 10
lightning days( corresponding to an isocerannic level of 47 lightning
days) we obtain 6,2 strokes per year, the same order of magnitude as
the allAre state uilding in "ew York and approximately 25-75 times
greater than on a 70-metre-high tower on level ground (10).
F.ow a leader develops near the ground is not known in detail. To
calculate the field strength on the ground near the leader certain
asnmptions have to be made as to th2 charge distribution in the leader
channel. Schonlaed assumed that the charges are evenly distributed
over the lmath of the lender (5). Iruec and Gold� (11), on the other
hnnd, assumed that the charges decrease exponentially with height,
covresponding to e0. The constant 110 is between 54 m and 1430 m.
Pierce (12) points out that the line density of charge along the channel
cannot be uniform. It is probably greatest towards the middle portions.
Tn the imnediate vicinity of the cloud the charges are slight, as the
aotential difference is comparatively small. At the ground end the
notential difference is great, but the time in not sufficient for the
extensive rroduction of charge. Grincom (13) evolvcd 8 theory according
to which tee annoy charea:s at the end of the !ender are concentrated in
a ball with n raTatively large diameter. The reasoijar his discusions
was an unexpectedly large number of flashovers in a high-tension network.
Pig. 2 shows the charge distribution according to these different
assumptions. In calculating the attraction distance between the leader
and an object on the ground, the charge distribution in Pig. 2 plays
an essential eart. Thus Golde (14) calculates the distance between the
leader, according to Fig. 2b,and the ground with different intcesity
values of the space charge for one field strenpth on the ground. A
leader charge of 1 Coul, according to Goldc, corresponds to a lightning
current of 20 kA. Distributed exeonentially, with hip = 1000 m, this
charee at a distance of 17 m from the ground generates a field strength
of 10 kV/cm. 'eith a latral distance of 15 m, calculated in the same
way, the field strength is 3 kV/cm and conseanently sufficient to start
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
t
U
5
fi
11.
-5�
a capture discharge from an object about 20 m high to the leader. The 1
/met/lode/ 4
attraction distance is thus 1172 + 452 = 48.1 in. Thislcalculation of the
attraction distance for smaller space charges and consequently smaller
current strengths receives support from many observed lightning strokes
P
on lower objects adjacent to high objects or horizontally to low objects.
Pig. 3 shows an example (15), a stroke horizontally on n farmhouse roof 4
to the lightni:ig-conductor cable, which was located the width of two
4
tricks from the edge of the roof. The damage was comparatively slight,
a fact which is often observed in connection with such nhenomena, for
example, in the first "classical" lightning stroke at Purfleet, England,
in 1777 (16) on the Meeting House of the Artillery College, which was
furnished with a lightning arrester. The lightning did not strike the
rod but struck an iron corner clamp above the cutlern. The tip of the
rod formed a protective angle of 31� and was 14 in from the point struck.
Here also the damage was insignificant, but the fact that a lightning
conductor with a protective angle of 310 could not prevent it caused a
rreat sensation at the time, without it being possible to give any
explanation.
The protection of high-tension lines against direct lightning strokes
with the aid of earth wires is an important technical problem which has
given rise to many investigations, both theoretical and experimental.
Davis (17) took up afresh the problem of calculating the protective value
of these earth wires and consequently the frequency of lightning strokes
to high-tension lines. He calculated the flashover voltage between the
end of the leader and the ground with the aid of an impulse flashover
gradient based on extrapolated experimental values. He determined by
geometrical calculations the effectiveness of the ground wires' shielding
angle in relation to the high-tension line. The shielding angles ranged
from 450 to 150. With a shielding angle of 450 in an earth wire at a
height of 20 in, flashes with a current strength of over 37 kA would be
intercepted by the earth wire. With a shielding angle of 20� the corres-
nonding current strength is 20 kA. It can be deduced from the statis-
1
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
� �
tical distribution of lightning currents that the number of lightning
strokes to the tranamission line would be reduced from about 40% to 175
(10 to 4.25). This does not tally with practical experience. The quan-
titative calculation is dependent on very approximate assumptions as to
the distribution of the charges in the leader. The geometrical method
cannot take into account metre-long corona discharges which are generated
from the cables and alter the geometrical picture.
Grindley (15) calculated the magnitude of the charges bound on earth
wire and on phase conductor by the leader in some distant part of the
wires. Equality of charges is taken as a condition in which both wires
are equally likely to be struck. The charges are calculated from the
field due to the leader and it is shown that equality of charges corres-
nonds to both wires being at the same equipotential of the leader.
Calculation sugrests that shielding is norAally adequate for conductors
in a wedge, of which the apex line is the earth and of semi-vertical
angle 45�. At a shielding angle of 45� an earth wire would consequently
offer perfect protection. Fig. 4 shows that this does not agree with
experience. This figure is a collocation of lightning faults in high-
-tension lines operating at 275-400 kV and a few line+elow 275 kV with
n low earth resistance in the pylons. The majority of these faults
arose from liehtnieg strokes to the line. The valnes ore derived from
n tabulation by Knoteoko (19) nod hove been converted to a line length
of 100 km, 100 lightning days and a line height of 30 m. According to
this tabulation, the faults are reduced from about 10 to 0.25 when the
shielding ang7e is reduced from 450 to 200. In addition to Kastenko's
values for protective lines about 30 m high, the values published by
Buresdorf (20) for 220 kV, 150 kV and 110 kV lines in the USSR are also
reproduced (Fig. 4). 'ihese lines have on average height of 15 m. The
original values have been converted from 30 lightning hours (. 20
lightning days) to 100 lightning days. The number of lightning strokes
to the phase conductor increases very substantially with increasing
shielding angle. The risk of a stroke increases approximately quadra-
pproved for Release: 2017/09/11 C06461858
- -Approved for Release: 2017/09/11 C06461858
- 7 -
tically with the height of the object (10). If these values, which were
estatlished for lines 15 m high, are converted to a height of 30 m
by a factor of 4, the agreement is then satisfactory.
T. Horvgth. (21) has attempted to avoid the disadvantages of a
geometrical calculation by means of an experimental model investigation.
determined the orobability distribution of the critical distance
between the lender end the grounded object with the guidance of Goldele
calculations nnd of the statistical distribution of lightning currents.
3y model experiments on a scale of 1:30 to 1:100 he determined the
probability of a lightning stroke on the conductor, which was protected
by an earth line with shielding angles of 20� and 30�. By reducing the
.sielding angle from 30� to 200, the number of lightning strokes was
reduced in the ratio of 10:1 in these model experiments on a scale of
1:50. This agrees with experience, as shown in Fig. 4. But the problem
is not thereby solved, as the experiments are based on uncertain assum-
ption S as to the critical distance. The relative amplitude of the pre-
-eir4charges in these model experiments is not in accord with reality.
These model experiments do not reproduce correctly the influence of
the heieht of the line on the frequency of lightning strokes. An
alteration in the scale of the model and thereby the height of the line
in the ratio of 2:1 results in a change in the stroke frequency in the
ratio of 30:1 with a shielding aegle of 200 and of 10:1 with a shielding
angle of 30�And therefore �of increement with observed vaives.
The question of the influence on the lightning path of corona
� discharges from lightning-conductor points is as old as lightning
research. This complex of questions includes Dauzere's inquiries Con-
cerning an accumulation of lightning strokes on a boundary line between
two different geological formations. According to Dauzere, the emanation
of radium from geological discontinuities influences the lightning path
through the ionization of the air. About ten investigations ensued.Sw*e
shoved a tendency to strive afterCiap-bly.The question would long ago
cemmoreca
(44% have been laid ad acta if substantial V40.�~4444, interests had not been
1
pproved for Release: 2017/09/11 C06461858
'Approved for Release: 2017/09/11 C06461858
8-
involved in protecting buildings centrally by a single radio-active
point. The le-radiation generated by these radio-active lightning con-
ductors can be demonstrated by sensitive instruments at a distance of
several hundred metres. But the ionization of the air should be 106 to
108 times more -powerful than the ionization that can be produced by a
radio-active point, to have any possible effect on the lightning path
(23). Uncertainty in judging electrical effects in th+tmosphere is
often a result of incorrect measurements. Unfortunately, many investi-
gations made before 1942 are valueless on account of fundamental errors
In measurement, particularly as regards the electrostatic field in
thunderstorms. In low-lying land with plants, trees and grass the field
is limited to .values which seldom exceed 10 kV/m. Incorrect measurements
at a height of 600 m above sea-level at the High Knob station in
, for example, resulted in field strengths with an average
value of 200 kV/m. Ten per cent of the measured values showed a field
strength of more than 260 kV/m (24). Even with a field of 6-10 kV/m
all plants produce such intense space charges that the electrostatic
field seldom exceeds 10 kV/m. With a field of 70 kV/m, discharges on
a rah's fingers are visible. Only in areas without these "points", for
example, at sea, can such powerful electric fields arise that St. Elmo's
fire can be formed on a ship. There is consequently a conceivable risk
of lightpng strokes on ships, but with the steel ships of today this
Is no longer of current interest.
The Lightning Path on the Ground
The statistical frequency of lightning currents and charges is
fairly well known through many investigations (25). Their probability
dietribution is represented within certain limits by a normal logarith-
mic distribution. Tf the percentual number of the mnenitude x (current
or charge) within the limits htx is denoted byaz, the distribution is
expressed by
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
LT; bie-1)
--9--
A, . �(xo) e-(log (x/x0))2/282 (1)
.1x =IT; Zan x
is value, M the base of the natural logarithm
(2.3Cr,) and s the standard deviation. The mode of distribution xmod is
_(,8)2
xmod = x0e -
ar,d the arithmetical average value xerith
xarith x0e
(2)
(3)
The nrotability of the magnitude x is calculated by the Gaussian distri-
bution
where
t J/
P -- - 1 r e-t 2i 2 elf
liTii oa
t = I log (x/x0).
S
(4)
(4a)
'-lome typical values are given in Table 1. Column 1 reproduces the
values veleIeetts recommended by the AIEE for lightning currents above
median
5 kA (26). According to this tabulation, with 18 kA as amczteg.z. value,
the current strength which occurs most frequently is 10 kA. The arith-
metical mean is calculated us 23.7 kA. Column 2 shows the average values
in investigations by both Berger and :_itekolnikov and measurements 1n
weden (25) which take into account current strengths below 5 kA. The
current strength which occurs most frequently is 4.25 kA. The arith-
metical mean is 22.8 kA. The charges in column 3 are valaes for a single
lightning stroke. Column 4 gives charges for a complete lightning dis-
charge consisting, of multiple discharges and prolonged discharges (25).
z]xtrepolation for a cumulati'Ve value less than 2% would give values
toe large for the current strength. If the percentage of lightning
currents greater than 50 kA is drawn on log- lin paper, 133 for J(A)
lightning currents, according to columns 1 and 2 in Table 1, is represented
by the following expressions:
P = e-o384 x 10-4(2 x 104 +
J1
P
0.288 x 104(3 x 104 + J). T e
u2
wth
J the current in A.
pproved for Release: 2017/09/11 C06461858
(5a)
(5b)
Approved for Release: 2017/09/11 C06461858
t
If the current strength is calculated for a probability P = 6.1
I 74bte
a
rTab0
3J
and less, aecording to (5a) and (5b), the values given in Table 2 are
obtained. Paecrful strokes with current strengths exceeding 200,000 A
occur very seldom.
)1 The ftwo**44 current-heat impulse i2dt has not been systematically,
investirated. Practical experience is available from investigations
with a power transmission line from 6oulder to Los Angeles (27) carrying
287 kV in an area with about 30 lightning days a year. Three stations
are shielded aeainst lightning strokes by capture towers 50 m high,
carrying earth wires. These towers and some pylons, 80 towers in all,
are furnished with elevation points, magnetic links for current measure-
ment and six coeper wires connected in series (see Table 3). The current-
-heat impulse values Oven have been calculated from experimental inves-
tigations byroitzik (28).
rs
'/,xperience over 20 years shows that, in lightning strokes on these
towers, 1, 2 or 3 wires are destroyed, never 4. One example mentioned
is a current strength of 43 kA, which destroyed wires nos. 1-3. On another
occasion 36 kA were measuredwithout wire no. 1 being destroyed. Bellaschi
(29) determined experimentally the connection between the current impulse
(decreasing exponentially) and the half-value period TH required for
fusing coeper wire:
.31 2 x 1 05 x A
(6)
where A is the cress-section in mm2 and TH the half-value period in /4s.
With a 36 A peak current which does not fuse 0.81 mm wire, the half-
-value neriod, according to (6), is less than 40 Ps. The probability of
these towers being struck
would seem to be at least 0.15 per year. With
80 towers and 20 years' experience the Probability of the current-heat
impulse being sufficient to destroy a 2.09 mm- copper wire is thus less
than
After 16 years of measuremenits at 3an Salvatore, Berger (9) confirms
stre
that the current-heat impulse was greater than 1.5 x 106 A2 on five
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
� 41 �
occasions and greater than 6.4 x 106 A2 a on one occasion. The corres-
:rad'
nonding -probabilities are about 0.75% and about 0.152 respectively.
These results are collated in Table 4.
The maximal velocity at which the lightning current increases in
- _
_ _
strokes to the ground (di/dt) has not been so comprehensively investi-
gated as-to enable the observational material to be treated statisti-
cally. Direct measurements on tall chimneys, carried out by Hylt4n-
-Cavallius and Strandberg (30), showed on several occasions di/dt
values greater than 30 Wes and once a value of about 100 kA/ms.
ti
At 30 klis the inductive voltage drop in a claim-1.e with an inductance
of 1.67,11/M is 50 kV/m. Oscillographic measurements of the lightning
current make an analysis possible. !-',erger (8) was able to show a
current rising to the maximum value after 5-10 las as a consequence of
a first downward-progress leader from a negative cloud and of an upward
midgap streamer from the air (Fig. 5a). Ten per cent of the measured
course shows a steepness greater than 25 kA/iPts. Partial strokes from
'negative clouds, which follow either the first discharge, in accordance
with Fig. 5a or the prolonged discharge typical of San Salvatore of a
few tens or hundreds of amperes, have a front period of only one or a
few microseconds (Pig. 5b). The specific steepness is thus substantially
(than with a current course in accordance with Pig. 5a. The method of
measuring the current course with the aid of a delay cable has been
Setrri
used since 1960. The revIlts are consequently not so numerous as to
(iaable
a statistical analysis to be carried out.
The multiplicity of courses which lightning current can take on and
in the ground results in numerous and varying phenomena connected with
questions of protection. l'xtreme and uncommon phenomena attract parti-
cular attention in this connection and are naturally more conspicuous
than the more common cases. The following phenomena are mentioned
briefly and are illustrated in individual cases by examples.
4fras' 1. Power phenomena, for example, in a lightning-conductor
- - - -
The power increases as the square Of the current strength. With a
d for Release 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
� 4.Z-
lightning current of 100 kA tnere COMC3 into existence a power of
� �
I
500 kp in n es44e '7:h1ch follows a
very pronounced cornice on a building.
Tn addition breaking Torces arise in straight lengths.
2. The following are some of the manifold voltage nhenomena at the
place of strikingiand starting from the point of striking:
(a) Voltages of up to several million volts as a result of the
resistance in transit to earth.
(b) Sliding discharges from the point of striking to the sur-
r-Avidings up to a distance of 50 in, arising in very poorly conducting
bedrock, for example, granite.
(c) Voltage differences in a cable as a result of rapid change
In the lightning current. The voltage difference may be 100 kV per
eigSe metre of at 60 kA///s.
(d)'"Displacement" of voltages to distances of many hundreds of
metres through fissures in rock (tunnel building), wire fences,
,eTa. power lines or underground cables.
:LoN4i
3. 'That phenomena.
(a) When the striking place consists of metal, power is supplied
to the surface of the metal equivalent to the strength of the
1ir7litning current times the anode voltage drop. At 100 kA a power
w)
of olmut 1000 kW is 'p1 led. The pcpr density in copper is about
BOO kW/cm- 11W In alumininM 350 kW/cm- in the I' trot 10 kts and then
eeeree cming to the reduction of the current density.
(b) in the interior of the metal current heat arises, which is
inversely pr000rtional to the fourth power of the linear dimensions
of the conductor, for example, the wire diameter or the plate thick-
ness.
(c) Transmision of heat from the lightning Channel to the sur-
roeedings, thereby causing fire. '1.-th a lightning stroke in sand
folfelrites may be formed. At 100 kA the power is approximately
150,000 r.7N/m in sand.
1. Pressure phenomena.
(a) in the rapid expansion of the lightning path as a result of
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
� 43 �
3
o rn)id o.rowth in the lightning current,
a pressure wave arises
which in rlenned Crom the lightning channel at supersonic speed.
The pasardammo pressure and the wave impulse may produce considerable
danage.
(b) The sale phenomenon may arise in the vaporization of metal
wires.
(e) Turbulent forces sla arise in the evaporation of water in a
Tres snlit in most cases. At 100 kA blocks of stone weighing
5 tons ua be torn off aliC rocks weighing 100 kg thrown 20 m.
This short ta'oulation will be illustrated by a few examples.
1. On June 15, 1956, lightning struck the church at Rudolzhofen
(Bayern, Germany) and did extensive damage to the lightning-conductor
inr,tallation and to the church (31). Fig. 6 shows a semi-diagrammatic
:)icture. From th top of the tower two copper conductors ran to earth, one
2
(4 x 3 = 12 wires, 30.6 mm) directly and the other (7 wires, 24.2 mm)
over the body of the church. The cables were torn to pieces in at least
seven places, denoted in Fig. 6 by th figures 1 to 7. (The conducter
with lightning faults 1 and 2 ran down the other side of the tower.
The power input similarly took place on the other side. Fig. 6 was'
drawn with this down lead on the front side for the sake of a more
perspicuous survey.) At 1 the slate roof was damaged. At the grounding
point 6 a hole appeared in the concrete pipe which had been used as a
mechanical protection. The rainpipe was damaged at I and TT. A flashover
had occurred from the down lead to the rainpipe at a distance of about
30 cm and from the rainpipe to the power cable, similarly at a distance
of 30 cm. The cable in the church (4 x 1.5 = 6 mm 2) was vaporized for a
length of 10 m, equal to 0.5 kg of copper. Considerable pressure damage
was done to the organ and the electrical installation. Examination of
the material showed that the temperature of the cable had been about
7c0�0. To heat copper cables of 54.8 mm2 to red heat, a current-heat
imoulse of 1000 x 106 A2 s is required and to vaporize 6 mm2 copper,
37 x 1dP A2 s. From these power and heat phenomena the strength of the
pproved for Release: 2017/09/11 C06461858
'Approved for Release: 2017/09/11 006461858
44. �
liehtning current can be estimated as 300,000 kA, which, with an effective
duration of at least 11 ms meant a cloud charge of over 330 Coul. Such
ehenoleena occur with a probability of less tnan 1:100,000 (see Table 2).
2. Pig. 7 snows the situa-.A.on in a lightning stroke which caused
censiderable damage to a house witnout the lightning conductor being
able to preven.,; it (32). The flash had struck a birch tree about 20 m
high at a oistance of 35 m from the house. The tree stood on granite cove-
red with a thin layer of soil. The area was strewn with large and small
stones. :Seveeal traces led from the tree, which was totally splintered.
One trace, more than 55 m long, led to the house. Several cubic metres
of earth ene etones hod been thrown up, six wiedows had been smashed and
40 or 50 litres of earth and stones had been flung into the upper floor
of the house. Via the cellar the flash had struck a wall socket for the
electric cable in the kitchen, about 1.5 m above ground level. The
cable (2 x 1.5 mm2) was vaeorized. The trace then disappeared into a
0
4 mm cablerthence to a water-pipe with a flashover approximately 10 cm
in length. Pour persons who on tnis occasion were sitting only 2 m away
from the vaporiz,d cable were uninjured. This phenomena - that a house
may be struck and damaged from under ground - is not particularly uncom-
mon in Scandinavia when granite is the bedrock (33). Experimental inves-
tigations of peak currents in clefts permit an extrapolation to a current
strengtn of 100 kA with a half-value period of 200 )us. In damp sand the
amount of energy developed in such cases is aperoximately 1000 kWs per
metre of the lightning path. This is equivalent to the energy developed
in the detonation of 215 g of dynamite or 350 g of gunpowder per metre.
N
Displacement of the voltage plays an Important technical part. Tn
tunnel construction in high mountains, primers prepared for blasting
have been exploded too soon and accidents have been caused. As a safety
measure the cables are provided with a metallic sneath (34) and special
primers are used which require substantially greater power for firing
than the ordinary primers, which ignite even with a power of 1 mWs (35).
Te Cables in the ground may be exposed to a direct lightning stroke at a
distance of several tens of metres from the place of striking. Telephone
pproved for Release: 2017/09/11 006461858
Approved for Release: 2017/09/11 C06461858
-- AS -
cables connected with overhead lines are exposed to indirect and direct
lightning action. Conious specialist literature shows the importance of
safety measures (36-40, a limited selection).
- 3. In a lightning stroke on a lightning conductor, high voltages
agiEc..1
may arise in the etkitig:, with the result that a flashover may occur indoors
to electric cables or telephone wires 40 or 50 cm away. any fires have
been caused in this way. These high voltages are then transferred via
power or telephone lines to adjacent houses. It is, nowever, often
possible to establish that the damage at a distance of about birom4 10 m
from the point of striking is fairly slight. McCarthy et al. (41) analyse
a lightning stroke on a church in north-western Pennsylvania with the
aid of installed oscillographs and magnetic links. The lightning stroke
shattered a 4" by" by 18' wooden rafter, left the steeple and terminated
on the wiring above the ceiling. The church had no lightning conductor -
the electrical installation was the lightning conductor. 3ome fuses were
blown and some lamps vaporized, but neither the watt-hour meter nor the
7.2 kV transformer 60 metres away was damap,ed or affected by the flash.
The lightning current was established as having been 31.7 kA. Of this,
4.0 kA rent to earth in the transformer, 19.5 kA to the high-voltage-
-grounded conductor ,ind 6.7 kA to the high-voltage phase conductor via
the transformer. the duration of this !!artial current was determined.
oscillocraphically as longer than 2000 pal. Similarly the damage to the
electrical system in the case of the powerful flash at Rudolzhofen (Fig.
6) was very slight. Pig. 8 shows the damage in the general plan, denoted
by figures 1-8; fuses blown, lamps vaporized and flashovers in junction
boxes. ,cp meter was dema:7ed, in spite of the fact that the distance to
some installations was less than 30 m from the place of striking.
4. Very high voltages may arise in lightning strokes on a thick bed
of sand, the foundation of which consists of a better conductor, for
exarlple, clay. The lihtoing channel goes almost perpendicularly down.
The voltage gred!ent is about 100-150 kV/m. After about 100/As of contact
with the lir:htning channel pressed into the sand, fuluurites arise. The
temnerature of the lightning channel, according to spectroscopic
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858 "
-/� �
measurements by r"andelstam (42), is 16,000-25,000�C. Fulgurites arise
at 1800�C. Their generation can only be established in a chance direct
observation of a lightning stroke in send (43). In the atmosphere the
nressure is nrocsgated as a detonation wave, which can shatter window
pones at a distance of about 10 m. Pig. 9a shows a remarkable case of
damace causer, by pressure to a window pane. An almost circular disc
(160/170 mm in diameter, glasa thic'cness 2.5 mm) with sharp edges had
been egt out of the pane. The window was an external one, separated from
the widnor'e.1 inner one by a space (,r about 50 mm. The round disc had
ftller. down betwe:m the two windows. (The window is in the Institute's
Ls,rcnives, IniLiT.:ted by Professor H. Norinderj bockmann (44) describes
similar damage (Pig. 9b) in June 1754 to a hothouse in Larlsruhe. The
window was torn out of the frames in which it was nailed and thrown out
into the r7arOen - a r-sult of the negative preuaure. The explanation of
the damne may be that the stress on the glass through the pressure wave
depreSSion
r..zsite=p:Et=r-e ave of the liOttning channel was reinforced to
pont tlrough the wave's eing propagated in the glass and
reficted at the edge of the glass.
ightning Protection in the Light of 3tandard Codes
The %now1e97e 7:10. .cuicLice of that time was summsvized by the T.ightning
Pod CrnIre"encc in 'ondon in 1H7d (45). About 1::5 years later individual
coluitrien - the 13A and c;ermany - ber.an to draw up instructions
and Fniding principles Oesling with the ,,rotecti.)n of dif-erent kinds of
buildings, towers, chimneys, ships and last but not least structures
containing inflammable liquids, gases and explosives. A tnbulation of the
codes svnilabe to me is given as ref. (46). in the whole, the instructions
r-re much the same but on closer scrutiny show diverment views which are
not due to the individual cho-octer of particular countries.
The -uetion of the zone of flroteetion is not Oalt with in a uniform
manner. 7n the UTiA (16, m) a nhelding angle of 45� in imnortant cases
and C3� 'n less important cases is consi0er,2d sufficient (7ig. 10).
ooroved for Release' 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
�
In -]ragland (46, 1) a shielding angle of 45� is given, but not, however,
for particularly important buildings, such as explosives factories, oil
sec! eetrel tanks, etc. In the 1J33R (46, n) the shielding angle is sub-
flivided. In a chimney 52 m high the shielding angle for the upper part is
7-ry limited (Fig. 11). For the lower part a protective radius of 40 m
fs stated on the basis of model experiments.
Contrary views on elevation rods are made clear by Figs. 12 and 13.
In the USA several elevation rods are recommended, for example, 15 on
"the tyoical installation on a barn group". In Germany special rods are
not recomended on the roof of a farmhouse. The roof conductors act 88 an
air terminal. Some ewe-inspiring objects (Fig. 14) included in the Code
for Protection against Lightning (1959) in the USA are probably of most
use on the psychological plane. According to British 3tandard 16446,1)
an air termination need not have more than one point and should be at
least 1 foot above the salient point on which it is fixed. We accordingly
see that the dimensions of the elevator rod, which was formerly 10 feet
high or more, are now only rudimentary (Fig. 15).
In all the instructions the question of reliable grounding plays an
important part. In :..:nglend a maximum value of 10 ohms is prescribed
eeohomlicoMy
(45, 1, 308c). Tn Austria this is not possible at ti-44440/-4-41$ justifiable
expense in certain provinces. The Austrian instructions (46, a, 10, 3)
resistAnce
the.refore allow higher timange***Pai values than 10 ohms for a specific
earth resistance of 'ore than 250 ohms/m. For this, exact instructions
with vaelon5 examples are given. Tn the Uir.iA (46,m, 2171i) it is laid down
that low resistance is. of course, desirable but not essential. By a buil-
ding resting on a base of solid rock it would be impossible to make a
E-round connection in the ordinary sense of the term. The most effective
71eans would be an extensive wire network laid on the surface of the rock
s..rrounding the building, after the manner of a counterpoise to a radio
netenes. ,.T.ere we aperoacn the standpoint of James Clerk Maxwell (47) that
"eerth" doe e not exist in the protection of buildings against lightning.
The .?ssent;a1 thing is to prevent potential differences. Wixwell in 1876
suggee�ted a construction like a cage with 6 mm2 copper wires. In
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
- 48 -
modern houses .t.h heating sy, terns, water-pipes and electric cables, not
many wires would te needed to complete the whole arrangement according
to 7h=e11's su-restion (Fig. 16).
The crosection of the conductor necessary for the discharge has
for easy :ecodes been fixed more on the 'basis of the craftsman's experi-
ence of muchanica7 strenp"th a::d of practical knowledge of very powerful
lir!Itnirg strokes than on the basis of physical investigations and
calculations of probabilities. For the protection of aircraft, however,
it '.as necesary to determine the cross-section of the conductor for
boeing par,oses. The current-carrying capacity of the bonding system
has to be such that a lightning discharge current can be carried between
eLy two extremities of the airplane without risk of damaging flight con-
trols and external surfaces or of producing excessive voltages within
the alveraft. The investiFations wore carried out with peak currents up
to 1r0 ik, renehinv the crest value at 10/As nnd dr')pping to 50 kA at
20/4.s. For this a copper cable with a cross-section of 3.3 mm2 or an
r,Thr.inium cable with a cross-section of 5.1 mm2 was necessary (48).
few exceptions the cross-section prescribed in the standards of
L_and is much Pigier,
dir!'erent crountries is a result of ,::)erience in building technique7
71,7,1.1.g protective systems can be divided into three groups: (1)
sup,r-irstullations, in which 1.1atively laro: sums of money have to be
:Tent to obtain perfect proteCtion: (2) standard .inntallations, which
buildiNfD
1,.re designed, in the first place, for valuable huilding-sOhd buildings
in which financial considerations play a small part; (3) "do-it-yourself"
installations, in which the financial aspect of protection plays the main
part and which are designed for the numerous small dwellings in the
provinces, for which a standard installation would be too expensive.
An example of a super-installation is shown in Fig. 17, an explo-
sives factory built on poorly conducting ground. The surroundings of the
building are protected by banks of earth. On these banks stand wooden
posts carrying a network of 50 mm2 copper wires with a mesh width of
about 8 m. This network is grounded through a ring conductor with out-
srea.t
going earth wires at a relatively simaLl distance from the building.
pproved for Release. 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858"
� 4 9 -
Thln In eonetructed like a Faraday cage and grounded separately. All the
metal ports In the building are carefully grounded in order to avoid the
aenerotion of small sparks 65 a lightning stroke. There is no permanent
metallic connection between the building and its surroundings. If the
of electricity is necessary, this is done by means of a cable, the
last 10 in of which are arranged overhead and are separated from the
building at this distance when lightning is forecast.
Standard installations are so thoroughly described in the instructions
of the respective countries (46) that it is not necessary to give an
account of them. Only one point may be mentioned: the practical difficulty
of connecting large metal parts to the lightning protective'aystem. A .
tniri
relatively large amount of damage has been brought about by, amongst
other things, television antennae. The difficulty is that in many cases
the supports for the antennae, usually iron pipes affixed to the roof or
a chimney, are not erounded at all. Obviously they are virtually ungrounded
lightning rods with only the twin lead of small wires as a cireuil; to
ground, either through a smell arrester, probably poorly grounded, or
through the TV set. The twin lead is easily vaporized and thereby pro-
duces an explosion (49).
In the majority of countrien lightning protective systems are not
economical for small houses: other considerations ploy the main part in
instanat4m.
their patadIMININNINI. Insurance statistics show that lightning damage in the
countryside seldom exceeds n value of 3.i; per insured small dwelling and
year. As a rule, it is not possible to produce a standard lightning
protector for an economical sum of 20 x 3 = $60. In Poland Szpor has sug-
gested protecting small houses with 10 mm2 iron wire (50). Several hundred
thousand installations have yielded a surprisingly good result as regards
lightning (16). ore detniled investigations of current-heat impulses imssi
and their probability showed that 10 mm2 copper wires are completely ade-
(rate (51). The instellation is made cheaper not so much by reducing the
cross-section of the conductor but by the fact that it is possible to use
lighter fittings and brackets, which are available mass-produced. 'A is A
eossible to stretch the wires over the building oneself, without having
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
� -20 �
to depend on the experts who are required for erection work with the
heavier standard wires. It is requisite, however, that owners of small
dwel.ligs who wish to build a lightning conductor themlves should
receive the essential instructions. This system is permitted in Sweden
(46, i, 9).
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
References.
1. B.L. Goodlet Lightning,
Journal of the Inst.EC4Eng. 81 (1937) July
82 (1938) Febr.
2. K.B.Mc. Eachron Lightning protection since Franklins day.
Journal of the Franklin Institute 253 (1952)
441 - 470
3. I.B. Cohen Benjamin Franklins Experiments.
A new edition of Franklin� Experiments and
Observations on Electricity.
Cambridge Mass. Harvard Univ. Press 1941
4. T. Bergman On mbjeligheten at forekomma hekans ekadeliga
verkningar.
(On the possibility of preventing damaging effects
of lightning)
Kongl. Vetenskape Academie Stockholm 1764.
5. B.F.I. Schonland, D.B. Hodges, H. Collens
Progressive Lightning V.
Proc.Roy.Soc. A 166 (1938) 56-75
6. B.F.I. Schonland, D.I. Malan, H. Collens
Progressive Lightning VI.
Proc.Roy.Soc. A 168 (1938) 455-469
7. B.F.I. Schonland
Progressive Lightning IV.
Proc.Roy.Soc. A 164 (1938) 132 - 150
8. K. Berger Front time and current steepness of lightning
strokes to earth.
Proc.Internat.Conf.Gas Discharges and the
Electrioity Supply Industry 12.
9. K. Berger Gewitterforschung auf dew Monte San Salvatore,
ETZ (A) 82 (1961) 249-260
(Compl. by pers. information)
10. D. Miiller-Hillebrand On the frequency of lightning flashes to high
objects.
Telles 12 (1960) 444-449
11. C.E.R.Bruce, R.H. Golde The lightning discharge.
Journal of the Inst.E1.Eng. 88 (1941) 487-505
pproved for Release: 2017/09/11 C06461858
5ti
Approved for Release: 2017/09/11 C06461858
12. B.T. Pierce
13. S.R. Griscom
14. R.H. Golde
15. F. Schwenkhagen
16. D. Muller-Hillebrand
17. R. Davis
18. I.H. Gridley
19. EV. Bostonko
20. V.V. Burgsdorf
Some Topics in atmospheric electrioity.
Recent advances in atmospheric electricity
Pergamon Press 1958
The prestrike theory and other effects in the
lightning stroke.
Trans.Amer.Inst.Electr.Bng. 77 (1958) 919-931
The frequency of occurrence and the distribu-
tion of lightning flashes to transmission lines.
Trans.Amer.Inst.Electr.Eng. 64 (1945) 902-910
Blitaschaden trots Blitzsohuts.
Elektrotekniek 1959 Nr 18
Noormanns Periodleke Pere N.V. Den Haag (Holland)
The protection of houses - an historical review.
Journal of the Franklin Institute 274 (1962)
34-56
Frequency of Lightning Flashovers on overhead-lines.
Proo.Internat.Conf.Gas Discharges and the
Electricity Supply Industry
The shielding of overhead lines against lightning.
Proo.Inst.Electr.Eng. 107 A (1960) p. 325-335
Contribution to the Conference of Cigri Study
committ6 8 in Athen (1962). Not published.
Lightning protection of overhead transmission
lines and operating experience in the USSR.
Cigrti report 326/1958
21. T. Horvolth The probability theory of lightning protection.
(In hungarian language)
Elektrotecknika 55 (1962) 48-61
22. C. Dauzare, I. Bouget Influence de la constitution Oologique du viol
sur lee points du chute de in foadre.
C.H. /wad. So. 186 (1928) 1565
23. D. NUller-Hillebrand Beeinfluesung der Blitzbeln duroh radioaktive
Strahlen und durch Raumladungen.
E T Z A 83 (1962) 152-157
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 606461858
3.
24. L.M. Robertson, V.W. Lewis,
C.V. Fonet Lightning investigation at high altitudes in
Colorado.
Trane.Amer.Inet.Electr.Eng. 61 (1942) 201-208
25. D. Mliller-Hillebrand Zur Physik der Blitsentladung.
E T Z (A) 82 (1961) 232-249
26.
27.a) B. Cozsens
A method of estimating lightning performance
of transmission lines.
Trans.Amer.Inet.Electr.Eng. 69 (1950) 1187-1195
Symposium on operation of the Boulder Dam
Tranomiesion Line-Insulation and Lightning-
Protection.
Trane.Amer.Electr.Eng. 58 (1939) 140-146
b) T.M. Bakeelee, E.L. Kanouee
Thirteen years lightning performance of Boulder
287 kV transmission lines.
Trans.Amer.Electr.Eng. 69 (1950) 796
Results completed by personel information
28. R. Foitsik Versuche nit grossen Stosetr8men.
E T Z 60 (1939) 89-92, 128-133
29. P.L. Bellaschi Heavy surge currents-generation and measurement.
Trans.Amer.Electr.Eng. 53 (1934) 86-94
30. N. Hylt�n-Cavallius, Field measurements of lightning currents.
1. Strandberg
Elteknik (Stockholm) 2 (1960) 109-113
31. A. HiSel Blitzschaden an Kirchen.
E T Z (a) 82 (1961) 288-293
32. D. MUller-Hillebrand iskrisk ooh Askekydd.(Lightning danger and
lightning protection)
Teknisk Tidskrift Stockholm (1960) 625-630
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 006461858
4.
�
33. D. Muller-Eillebrand
34.. K. Berger
Rackvartiger Blitzeinschlag in Hauser und
Energieumsatz im eingeengten Blitzkanal.
E T Z (A) 78 (1957) 548-553
Notwendigkeit und Schutzvert metallischer Mantel
von Sekundarkabeln in Hochetspannungsanlagen
und in Hochgebirgestollen ale Beispiel der
Schutzvirkung allgemeiner Faradaykafige.
Bull. S.E.V. 51 (1960) 549-563
35. K. Berger, I.P. Fourestier,
H.F. Schwenkhagen Blitzschutz far elektrieohe SprengzUnder im
Stollenbau.
Nobelhefte 25 (1959) 149-160
36. B.L. Coleman The direct lightning stroke to a buried cable.
The Electrical Research Association
Leatherhead, Surrey
Technical Report 071 (1951)
37. H. Meieter Blitzschutz an Telephonanlagen.
Technisohe Nitteilungen P T T
36 (1958) 13-32
38. D.W. Bodle, P.A. Gresh Lightning surgee in paired telephon oable facilities
The Boll System
Technical Journal 40 (1961) 547-576
39. S.I. Little Protection of exchange equipment and subecribers
installations from damage due to lightning and
contacts with power lines.
The Poet Office
Electr.Eng.Journal London 53 (1961) 219-225
40. E. Foretay, R. Ruchet Schutz von Kabeln in Wasseretollen gegen Blitz-
sohaden .
Bull S.E.V. 52 (1961) 33-39
41. D.D. Mac Carthy, D.R. Edge,
D.A. Stann, W.C. Mc Kinley Lightning Investigation on a rural distribution
System .
Trans.Amer.Inst.E1.Engs. 68 (1949) 428-437
pproved for Release: 2017/09/11 006461858
Approved for Release: 2017/09/11 C06461858
5.
42. S. Mandeletam
43. M.N. Reed
Tiber die Temperatur des Blitzes und die
Sthrke des Dormers.
Proc. Ionization phenomena in gases.
Edited by H. Maecker 1962 North-Holland Publ. Co
Amsterdam
Fulgurites in the making.
Rocks and Minerals 33 (1958) 406
Earlier references ins
J.I. Petty. The origin and oocurence of Fulgu-
rites in the Atlantic Coastal Plain.
American Journal of Science V, 31 (1936) 188-198
(86 references)
A.F. Rogers. Sand fulgurites with enclosed Lechate-
lierite from riverside country, California.
The Journal of Geology 54 (1946) 117-122
(15 references)
44. J.L. Bookmann Caber Blitzableiter.
Karlsruhe 1. Aufl. 1783. 2. Aufl. 1830
45. C.I. Symons
46.a) Austria
b) Czekoelovakia
c) Denmark
Lightning Rod Conference.
Report of the delegates. London 1882
see alsot Engineering 1882 p. 225
Leitshtze fUr die Errichtung und tberprUfung von
Blitzschutzanlagen.
Wien (1960)
Elektrotechnischer Verein Osterreiche
Instruction for the installation of lightning
proteotive systems.
Prague (1955)
Vejledende Regler .or Ud4relse at' Lynaflederanlaeg
paa Bygninger m.m.
14benhavn Valby (1944)
Elektrioiteteraadet
pproved for Release: 2017/09/11 C06461858
-
SWAYS �������������...., ,,,
Approved for Release: 2017/09/11 C06461858
6.
46.d) Finland
e) Germany
0 Netherlands
lekskydd for byggnader.
Beleingfors (1943)
Blektrieka Inepektoratet
Blitzsohutz..
Berlin (1957)
Auseohuss fUr Blitzableiterbau e.V.
Richtlinjen voor Blikeemafleiderinstallaties.
NEN 1014
Den Haag (1958)
Nederlande Elektrotechnisoh ComitS
g) Norway Lynavlederboka.
(1951)
Norsk Brannvern Forening
h) Polon
i) Sweden
Protection of building against atmospheric dischar-
ges. General instructions.
Warsaw (1955)
Byggnadelekledare.
SEX Handbok 2
(1960)
Sveriges Standardiseringekommission
k) Switzerland Leiteatze fUr Blitzschutzanlagen
(1959)
Sohweizerisoher Elektroteohnischor Verein, Ulrich
1) U.K. Protection of structures against lightning.
(1948)
British standard code of practice CP 326.101
m) U. S. A. Code for Protection Against Lightning.
(1959)
National fire proteotion assooiation
n) U.S.S.R.. Lightning proteotion for industrial and other
buildings.
Moskva (1951)
I.S. Stekolnikov, V.S. Komelkov, A.F. Bogomolov,
P.A. Lihachev, V.N. Borisov, L.N. Lopshitz
pproved for Release: 2017/09/11 C06461858
t�--, r
Approved for Release: 2017/09/11 C06461858
.47. 3.0. Maxwell' -
07.
On the proteotion of buildings from lightning
Rep.Brit.Assoo. for the advanosmentvsoienes
(1876) p. 45
48. A.O. Kemppainen,,I.H. Merriman
� 49. S. Beck
50. St. Szpor
Aircraft bonding for lightning proteotion.
Proo. 1948 Symposium Lightning protsotion for
aircraft.
Lightning and Traneiente Research Institute
Minneapolis, Minnesota
Westinghouse Bl.Co, East Pittsburgh personal
information.
Paratonnitres rureaux de type 16ger.
Rev.Osn. de liElectrioit6 68 (1959) 263
51. D. K011sr-Billebrand Vber d41 Beanspruchung und Bemeseung von
Blitzschutzenlagen.
E.u.N. 77 (1960) 345-349
4
pproved for Release: 2017/09/11 C06461858
'
�
Approved for Release: 2017/09/11 006461858
Tables
Table 1. Normal log distribution.
,
Column
1 1 2
3 1 4
Magnitude
Current, kA
.Charge, Coul
mwatageftleclia..ki
18
13
3.1
15
Deviation
0.32
0.46
0.40
0.51
'.ode
10.4
4.25
1.33
3.8
Arithmetical mean
23.7
22.8
� 4.7
30
Table 2. Extreme curvent strengths.
Current 1, A
Current 2, A
.Probability
1:10
42,000
50,000
d:100
100,000
130,000
d:d000
160,000
210,000
1:10000
220,000
290,000
Table 3. Current heat impulse
No.
Wire diameter,
:TIM
Wirecnmstaion
2
MI
Current heat impulse,
106 A2Stc
1
0.81
0.52
0.019
2
1.02
0.82
0.048
3
1.29
1.31
0.12
4
1.63 �
2.09
0.303
.
.
5
1.83
2.63
0.49
6
2.05
3.30
0.77
Table 4. Probability of current heat impulses.
Region
Probability
C'irrent heat impulse
106 A2setc
Zoulder
1. Significance tests have also been carried out between
/D
the mean values of the dispersion ratio for the various intensity
groups. Significance was calculated tobeat the 0.04 level for groups
(a) and (b) and at the 0.29 level for groups (b) and (c).
It is concluded, therefore, that the dispersion of the second
component of a two component whistler is about 3% greater than that
of the first and that there are significant differences between
the dispersion ratios when two component whistlers are grouped
accordingly to the relative intensity of the two components.
Whistlers with more than two components.
A similar analysis of the dispersion ratio for successive com-
ponents in the case of whistlers with more than two components, leads
to a similar conclusion. Results for records of whistlers with three,
four and five components are shown in Table 2.
No. of
records.
Mean value of
Dispersion Ratio.
Significance
level P.
52
2/D
1.023
+ 0.007
0.005
1
52
�
1.031
+ 0.007
0.001
3/D2
20
D4/D
1.018
+ 0.015
0.30
3
7
D5/D4
1.011
+ 0.020
0.70
Table 2. Ratio of dispersions of successive components
for whistlers with more than two components.
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
5.
Although the dispersion ratios D
4/D7 and D5/D are both greater
than one the number of available records' is 4
too small for any great reliance to be placed on these figures.
The Relationship between Dispersion and the Time Interval
between Components.
An investigation was carried out to see whether the difference
in dispersion between successive components was in any way related
to the time interval between components. The analysis was made by
plotting LID, the difference in dispersion between successive com-
ponents, against n t, the time interval between components. Time
intervals of 20 ms were used and the results were obtained from all
records of D2 - D1 and D3 - D2. In both cases it was found that
there was an initial increase Intl D, as d t increased, followed
by a decrease. The maximum value for 6 D occurred for a value of
a t 30 me.
The results for D2 - D1 are shown in figure 3, the numbers in
brackets indicating the number of records, in the particular time
interval, available for calculation.
20 40 60 BO 100 120
(nsec)
Fig. 3. Relationship between the difference in
dispersion between the first two whistler
components and the time interval between
them.
Speculations on a Possible Explanation of these Results.
From the above it would appear reasonable to assume that (a) in
the majority of cases the source of a whistling atmospherio is a
lightning discharge between clouo and ground;
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
6.
(b) in multiple flaeh whiutlere the several components
arise from the separate strokes of the lightning
discharge,
(c) there is a small increase in dispersion between the
successive components of a multiple flash whistler;
(d) the increase in dispersion between whistler coaponente
is related to the time interval between them.
Point (d) is of particular interest as it suggests that
the dispersion measurements may be related to the physical
processes occurring in the thundercloud during the interval
between strokes. A possible explanation of the variation in
D might be sought in terms of these processes.
Factors Influencing the Dispersion of a Whistler.
If
by D .. 1� �2 de.
The dispersion D of a whistler is given
(3torey,1953), provided the frequency of the whistler is much less
than either the gyrofrequeacy fH or the plasma frequency fo.
de is an element of path length and the integration is taken
over the whole path. By substituting for fo and fH,
D .. ( (N.I de. where e is the electronic charge in e.e.u.,
�)
P i
N the e ec run density per cc and H themagnetic field strength
r
in oersted.
It is known that the ducting of whistlers, giving rise to
the multiple path type, occurs very much more frequently in
higher latitudes than it does at lower latitudes. It would
seem reasonable, therefore, that the energy in each component of
a multiple flash type has traversed the same path. In the short
time interval between components the values of the magnetic
field strength along the path will not change. Consequently,
any change in dispersion between one component and another can
only be due to a change in electron density along the path.
The fact that /I D initially increases with d t, as shown
in Fig.3, suggests that the change in dispersion is related in
some way to the physical processes occurring in the thundercloud
during the time interval between strokes. Could then the source
of electrons, necessary to account for the increase in dispersion
originate in the thundercloud?
Runaway electrons were postulated by ....P.R. Wilson (1925)
who showed that in the presence of electric fields such as are
found in thunderclouds, the energies of such electrons could
be as high as 109Nev. Evidence for such penetrating particles
of high energy has been found by ..,nonland and Viljoen (1933),
using geiger counters and by Halliday (1941) using an expansion
chamber. In the former case there was a pronounced tendency
for the counting rate to increase at the moment of the flash and
in addition more impulses were registered during the few seconds
i �Droved for Release' 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
7.
before a flash than during-a similar interval after the flash. The
radar studies of Atlas (1958), Hewitt (1957) and Rumi(1957) have
shown the existence of upward ionised jets during discharges of a
thundercloud to ground. Such jets extend to heights well above the
top of the thundercloud and are estimated by Rumi to have a velocity
of approximately 2 x 107 cm/sec.
Using data given by Nelms (1956) for the range of electrons in
various media it is estimated that for an electron to penetrate the
remaining atmosphere from a height of 10 Km. it must have an energy
of the order of several hundred Mev. It is possible therefore that
runaway electrons, accelerated within the thundercloud, could provide
an upward jet of electrons.
Assume, in the first place, that such a jet of electrons moves
upwards during the time between one stroke and the next. The modal
value of 4 t between first and second strokes is 40 ms. and for this
time interval 6 D 2.0 sec*. If the upward velocity of the jet is
assumed to be 2 x 107 om/sec.)in 40 ms. an ionised column 8 Km long
will be formed. The electromagnetic energy radiated from the second
stroke would thus pass through this charged column in addition to
traversing the whistler path traversed by the energy from the firat
stroke.
The question now arises as to whether this column of enhanced
ionisation can satisfactorily account for the increase in dispersion of
the second component over that of the first. Assuming a magnetic field
strength of 0.12 oereted the value of the eleotron density in such a
column, which could account for an increase in dispersion of 2.0 seo
is 9.7 I 107 electrons/co. For suoh a medium the quasi-longitudinal
approximation of the magneto-ionic theory, upon which the expression
for D is based, is applicable and, assuming a colision frequency of
10/
10 /sec., the medium would have a refractive index of 22 for a 5 Kcisec
wave.
The above value for the electron density necessary to account for
the measured dispersion is the effective density required in the
assumed column. It could be used to estimate the current density in
the jet and the total charge moving upwards from the thundercloud. In
this case, however, values obtained would be minimum values for two
reasons. Firstly the picture of the upward moving column has been
greatly over simplified and no account has been taken of the effect
of recombination. This would increase the value of the electron
density by at least an order of magnitudeFrom the intensity of radar
reflections the electron density in upward jets has been estimated
at 5 x 10 per cc. as it leaves the cloud and 2.8 x 107 per cc. at a
height of 60 Km. These figures are consistent with the known values
for recombination coefficients at these heights which are of the order
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
V
8.
of 10-7. The calculated value for the electron density necessary to
account for increased dispersion falls satisfactorily within this
experimental range of densities.
Secondly it has been assumed that the jet current flows for the
whole of the time interval between strokes. From the work of Schonland
and Viljoen (1933) this assumption appears to be justified, but it
is likely that the intensity of the upward jet increases to a maxi-
mum at the time of the discharge.
The current density, r , in the jet is given by J Nev.
and using the calculated value of N 9.7 x 107 per cc. and
2 x 107 cm./sec. is equal to 0.31 ma/om2.
The total charge carried upward by the jet is given by
Q A.L.N e. when A and L are the area of cross section and length
of the ionised column respectively. Substituting known values
Q A. coulombs with A in square centimetres. Estimates of the
1(15'
radius 5f upward jets vary over a wide range from a few centimetres
to several kilometres depending on the assumed model of the thunder-
cloud. With the uncertainty in the value of A it is difficult to
make an estimate of Q. However, assuming a radius of 1 metre,
approximately 0.7 coulombe of charge would be carried upward by the
Aa-.1
!
jet. Increasing the radius soon leads to enormous values for the
charge carried upwards and unless the radius of an upward jet is of
the order of 1 metre or less the assumptions made above become
untenable bearing in mind that the charge brought to ground is
approximately 4 coulombs.
The effect of upward jets on field charge measurements.
If upward jets of electrons as postulated, do in fact exist
some evidence for them might be expected from field change studies of
lightning discharges. Malan and Schonland (1951 A) have considered
the electrostatic field which would be produced by an upward moving
charge and have shown that there is a reversal in the sign of the
measured field change as the charge passes through a reversal height.
Assuming that measurements are made at a distance D from a vertical
discharge, the reversal height, Hr, is given by Hr D/p.
For
Bees upward moving positive charge, electrostatic field changes
would be positive whilst the charge was below the reversal height and
negative when above this height. In the case of upward moving electrons
the signs of the field changes would be reversed.
Malan and Schonland eeplain th, observed results in terms of an
upward moving positive junction streamer between strokes pit having a
velocity of approximately 3 x 106 cm.seo. Final slow positive field
changes observed for flashes at a considerable distance may also be
pproved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
9.
due to positive streamers from the top of the thunderoloud.
Consider the simultaneous existence of upward moving junction
streamers of positive charge and upward jets of electrons. If both
processes are below Hr the electrostatic field changes produced by
them would be of opposite sign. Since both streamers originate in
the thundercloud the faster moving electron jet would be the first
to pass through the reversal height. When this happens the field
changes due to each process would be positive and a sudden increase
in the rate of change of the electrostatic field might be expected.
Such field changes are occasionally observed in the final field
charges of fairly distant discharges. A typical example of such
a field change is shown in Fig. 4.
Electrostatic
Field
'C---..........1ncrectse in rote
of change of field
Time
fig,A. The electrostatic field of a fairly distant
discharge showing a sudden increase in the
rate of change of field during the final field
change.
suppose now that the following information is availables
Distance of the discharge from the observer and hence Hr; the
th
the
jet,
to it
ight of origin of the final stroke; the time interval, t, between
e final stroke and the increased rate of change of the field. From
Be data it is simple to estimate the velocity of the upward electron
assuming that the increased rate of field change may be attributed
s passing through the reversal height.
he results of these calculations are shown in Table 2. for
nine re
Price In
stroke he
ords generously provided by Dr. D.J. Malan, of the Bernard
stitute, from data collected by him over many years. The
ights used in the calculations have been taken from Malan
and Schonl
change occu
reversal flea
and (1951 B). In all cases the increased rate of field
rred after the final strokr which originated below the
ght.
pproved
for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
10.
Record Distance H Field chaage Assumed t Velocity
No. (Km) (KM). increase after height (me) of elec-
stroke No. of die- tron jet.
charge (cm/sec x
(Km) 10-7).
Dk1,3
30
21.2
3
5.4
65
2.4
DAE3,1
20
14.1
2
5.1
14
6.4
DE2,3
16
11.3
3
5.4
20
2.9
DAH1,1
15
10.6
2
5.1
32
1.7
DAB5,9
15
10.6
3
5.4
60
0.9
DX4,6
10-20
10.6
9
8.9
28
0.6
D03,5
13
9.2
2
5.1
16
2.6
DE4,6
12
8.5
2
5.1
46
0.7
D02,3
10
7.1
1
3.7
.50
0.7
The calculated values of the electron jet velocity lie within
the range 0.6 x 107 - 6.4 x 107 cm/sec. Hearing in mind that there
could be a considerable difference between the assumed and actual
height of the final discharge, these estimated values of the velocity
may be regarded as consistent with those obtained from radar studies.
The results of the above paragraphs appear to lend support to
the assumptions made. In spite of this, the suggestion that the
increased rate of field change is due to the processes outlined
must be regarded as tentative for several reasons. The 9 records
analysed were the only ones showing the effect out of a total of 285
records and a more frequent occurrence of the effect might be expected.
Similar increases in the rate of field change between strokes earlier
than the last might also be expected but there is little evidence
for this. Finally, out of a total of 159 field chat3e records of
flashes which occurred at a distance of 8 km. or nearer there are
three instances of an increase in the rate of field charge occurring
after the final stroke. These results cannot, however, be explained
in a similar fashion as it is likely that all processes took place
above the reversal height.
In spite of these difficulties it is felt that further radar
studies, designed specifically to investigate the existence of ionised
jets above thunderclouds during the intervals between strokes, would
be justified.
ACKN0WL7DGEMENT.
The interest shown by the members of staff of the Department of
Physics, at the University of Natal, and their assistance with cal-
culations is appreciated.
Approved for Release: 2017/09/11 C06461858
Approved for Release: 2017/09/11 C06461858
BIBLIOGRAPHY.
Atlas, D. 1958 Recent Advances in Atmospheric Elec-
tricity. Pergamon Press, 441.
Bruce, C.E.R.
and Gold,R.H. 1941
Halliday,E.C. 1941
Helliwell,R.A.
and Morgan,M.G.1959
Helliwell,R.A.
Taylor,W.L., and
Jeans,A.G. 1958
J.I.E.E. 88tii) 487.
Phys. Rev. 60, 101.
Proc.I.R.E. Az, 200.
Proc.I.R.E. 4, 1760:
Hewitt,F.J. 1957. Proc. Phys. Soc. 212, 961.
Iwai,A. and
Outsu J.
Kitigawa,N. and
1956 rroc. Res.Inst.Atme., Nagoya University,
1, 29.
Kobayashi,M. 1958
Malan, D.J.
Malan,D.J.
Malan,D.J. and
Schonland, B.F.J.19511
Malan,D.J. and '
Schonland,B.F.J. 1951B
Nelms, A.T.
Recent Advances in Atmospheric Elec-
tricity. Pergamon Press. 485.
1955 Ann. Geophys. II, 427.
1958 Recent Advances in Atmospheric Elec-
tricity. Pergamon Press. 557.
P.R.S.A. 206, 145.
P.R.S.A. zu, 158.
1956 N.B.S. Ciro. 577, July, 28
Norinder,H and
Enudeen,E. 1961
Planet Space Science. .1,46.
Rusi,G.C. 1957. J. Geophys. Rea. 62, 547
Sobonland,B.F.J. 1956 Handbuch der Phyeik. 22, 578.
Schonland,B.F.J.
and Viljoen,J.P.I. 1953 P.R.S.A. 1A2, 314.
Storey,L.R.O.
Wileon,G.T.R.
1953 Phil.Trane. Roy.Soo. Aak, 113.
1925 Proc. Camb. Phil. Soc. 22, 534.
LIIIIIIIMIMINIMIEM1111Approved for Release: 2017/09/11 C06461858