Diurnal and seasonal variations of point discharge current during thunderstorms at a tropical inland station

Diurnal and seasonal variations of point discharge current during thunderstorms at a tropical inland station

S S Kandalgaonkar, M I R Tinmaker, M K Kulkarni & Asha Nath

Indian Institute of Tropical Meteorology, Pune 411 008

Received 3 September 2002; revised 2 December 2002; accepted 30 April 2003

The data of point discharge current (PDC) measurements during a total of 65 thunderstorms at Pune (18°32'N, 73°51 'E) are analyzed to study the PDC local diurnal variation and some issues related with the current. The analysis shows that about 83% of the total quarter hours occupied by PDC are localized between 1400 hrs L T and 2400 hrs LT and the remaining 17% are sparsely distributed over midnight to early hours in the morning. The net result of this study revealed that PDC is the dominant agent for the negative charging of the earth's surface and should be identified as an active element participating in global electrical circuit. The phase relationship between the positive and negative current during the diurnal period suggests that the active period of current of one polarity has a preferred time of occurrence over the other and by the late night hours the positive charge showed its sustained occurrence over the negative one. The seasonal relationship between storm averaged PDC and monthly mean maximum surface air temperature has also been examined. The comparison of seasonal average amplitude of PDC has suggested a strong positive correlation with the monthly mean maximum surface air temperature. The average amplitude of PDC during thunderstorm in premonsoon season is found to be about three times higher than those during the monsoon season. An examination of association between duration and amplitude variation of PDC showed that storms that are closer to the station within 4 km have longer duration and stronger current amplitude than when they are more than 4 km away.

Key words: Point discharge current, Thunderstroms, Atmospheric electricity

1 Introduction

Under the influence of high electric fields beneath storm clouds, the point discharge current (PDC) is known to flow through the pointed tips of elevated earthed conductors and topographic features of surface irregularities. It is explained that due to intense electric field at the tips of elevated points, electrons are accelerated to ionize air molecules in its vicinity and thereby produce a corona discharge. This phenomenon establishes the flow of a current that is known as PDC in atmospheric electricity (AE).

A large number of studies on the measurement of PDC during the early years and mid-fifties of the twentieth century and also in later years are available¹-14

•

The pioneering works¹'2

were with regard to the contribution of PDC to the global electrical balance sheet. This studies concluded that the contribution of PDC towards the maintenance of surplus of negative charges on the earth's surface is dominant. Chalmers ¹⁵ has provided a summary of the fundamental aspects of PDC and many other results relating PDC with electric field. These studies provide a remarkably complete account for most of the temporal and other features of PDC . It may be noted that a majority of these studies were carried out sporadically during time-bound field experiments to accomplish specific

objectives of the study, but the significant issue of the diurnal variation of PDC received a little attention¹⁶

., d. 17-19 F h d' 17-19 . .

except a 1ew stu 1es . rom t ese stu tes It IS

noticed that there exists a large discontinuity in the periods of observations. It is felt that specific information on the diurnal variation of PDC from another tropical location may be of vital importance, since electrification of deep convective thunderstorms in the tropics is believed to have a major role in the

I b I I . . . 16 20-23

s

f h d' g o a e ectnc ctrcmt ' . orne o t ese stu 1es have also shown significant differences in the dynamical and electrical properties of the convective cloud system of the tropics during premonsoon (break period) and monsoon season. Information on the diurnal variation of PDC is very scanty and has been a long standing requirement. In this paper, the data of PDC measurements during a total of 65 thunderstorms at Pune are analyzed to study the diurnal variation of PDC and some issues related with the current.

2 Experimental arrangement

A platinum/10% iridium needle constituted the conducting point discharge element in the measurements of PDC. The needle was erected on the top of the terrace of a three-storey building of the office of the India Meteorological Department (IMD)

222 ^{INDIA.N J} RADIO & SPACE PHYS, AUGUST 2003

at Pune. The height of the needle tip when installed was well above ail the natural or artificial objects in its vicinity. Details of measurements and some results of PDC studies are available elsewhere²⁴-27

. The signal of PDC received through the needle was termed as positive when the overhead net cloud charge was positive and vice-versa^15. The chart drive speed during the measurements was 1 em/min or higher (5, 10 em/min). One-minute interval time marks facility on the chart record, with a break at full hour, using a precision pendulum clock. A relay timer was used to monitor the time. These arrangements at the recording site have rendered convenience in the analysis of the PDC data in the desired manner. Pune and the surrounding region usually experience fair weather during November-March. During this period, air is dry and cold and the occurrence of thunderstorms is minimal. Weather during April-October is hot and wet. Usually, this is the major period during the year to experience thunderstorms²⁸'29

. The routine of the present observations of PDC during thunderstorms at Pune is necessary to mention.

The results of the aforesaid two studies, i.e. diurnal variation of thunderstorm and their frequency of occurrence over India, suggested that the most prominent diurnal period for thunderstorm occurrence at Pune was during 1500-2100 hrs LT, and their occurrence beyond late night and early morning hours was a very rare phenomenon. During the years, 1971- 1977, we were also not much aware of the many present day significant results ¹⁶ of PDC. Therefore, the present measurements of PDC commenced from the afternoon (AN) of those days which would give rise to thunderstorm and were continued tili the next day morning. This routine of observations was followed at the site from April to October during the

!Six-year period (1972-1977). All the PDC records thus obtained have been analyzed and studied.

It is realized that although the collection and study of PDC data during thunderstorms are important, it is just not sufficient. Non-thunderstorm episodes with appreciable foul weather electric field also contribute to the occurrence of PDC. The PDC data collected during thunderstorms alone do not allow us a strict comparison of the diurnal variation of PDC with global circuit response. This has been rightly pointed out by Williams and Heckman¹⁶ for their study of local diurnal variation of cloud electrification and the global diurnal variation of negative charge on the earth. Unfortunately, we do not have those records.

3 Data and method of analysis

The data of PDC measurements made during a total of 65 thunderstorms spread over the months April- October for the period 1971-1977, have been used to investigate the local diurnal variation of PDC. Table I shows the monthly number of storms used in this study. It is to be noted that for the year 1971, data for only one storm in the month of September were available. Although the data for this storm are included in this study, the results best represent the diurnal variations which are based on the storms over the six-year period (1972-1977). As already mentioned, the chart drive speed in the present set-up was 30 times faster than those of the observations in the past'⁸·'9

. It was, therefore, possible to make short- period time markings on the charge records and pick- up a large number of data points to adequately represent the current. lt is, therefore, planned to study the diurnal variation of PDC on the quarter hom (QHJ time-scale. To begin with, the PDC record on each day was marked with QH time markings (i.e. 0000- 0015 hrs LT = QHl, 0015-0030 hrs LT = QH2; 0030- 0045 hrs LT == QH3; 0045-0100 hrs LT = QH4 ... .

2300-2345 hrs L T= QH95, 2345-2400 hrs L T =

QH96). Thus, the total local diurnal period of 2-.J. hr

Table ]-Monthly number of thunderstorms during 1971-1977 at Pune used for PDC analysis

Monlhs No. of thunderstorms during the ~ears T01al

1971 1972 1973 1974 1975 1976 1977

March I 01

April 2 3 2 07

May 3 ^{.::; ~} 10 5 2 ^,~ -·'

JC~ne 8 6 7 21

July 2 02

August 2 02

September 02

Oc10ber 7 07

Total OJ 05 23 16 05 06 09 65

(LT) is represented by 96 QH intervals. Whenever PDC existed during a QH, current data ()..lA ±) were read from the chart and stored polarity-wise in that QH time-slot. The durations (m±) of these cunents, in minutes, were also noted.

At the end of the analysis, data accumulated on cunents ()..lA±) and their duration (m±) were added to obtain storm averaged cunent ()..lA±) and average duration (m±) for each QH interval. Using these data, the average charge (mC±) received by the earth's surface was computed employing the equation¹⁹, Q = I x t; where Q is charge in millicoulomb, I the cunent in microampere and t the time in seconds.

4 Results and discussion

· A general comment is made on the present result of the local diurnal variation of PDC shown in Fig. 1.

The PDC data conespond to the total 65 occasions of thunderstorms, occurring over a period of seven months (April-October), comprising the premonsoon and monsoon seasons at Pune. It is noted from Fig. 1 that the signals of currents in the morning hours are mostly absent, and the ones during the afternoon hours are stronger than those mentioned in the earlier analyses¹⁶-19

where the contributions of PDC from all foul weather episodes inclusive of thunderstorms over a complete annual cycle were discussed. The PDC data in this study correspond to the occasions of only thunderstorms. It is explained that the absence of the data in the morning hours is a characteristic feature of the local environment, and the strong signal during the afternoon hours is attributed to the sh01tcoming of the data. Probably, this is also the reason for the apparent noisy picture of the data. In view of the above explanations regarding the nature of the PDC data, it is submitted that the results of this study may be taken with some reservations.

4.1 The local diurnal variation of PDC and cha• ge

The storm averaged local diurnal variance of PDC ()..lA±) and charge (mC±) received by the earth's surface is shown at QH time-scale in Figs. [1(a)-(d)].

It can be noted from Fig. 1 that nearly 83% of the total QHs occupied by PDC are localized between 1400 hrs LT (QH56) and 2400 hrs LT (QH96); and only a small fraction (17%) of it is sparsely distributed over midnight (QH96) to early morning hours (0730 hrs LT, QH30). Manohar et al.²⁴ examined the timings of occurrences of spells of PDC for 10 and 11 premonsoon season thunderstorms of

the years 1987 and 1988, respectively, at Pune to understand the frequency of PDC occurrence during different diurnal hours. Their results showed that, on a storm day, about 80% of the total activity of PDC is usually accomplished prior to 2000 hrs LT, and the remainder is over by late night or early hours in the morning. Reports of the development of a thunderstorm during the early hours of a day at Pune have not come to the notice in the past, and this could be a very exceptional phenomenon. Th.e statistics of the PDC data (Fig. 1) indicated that the storm averaged maximum, mean and minimum cunents during the diurnal period were : -0.90; -0.47 and -0.14 )..lA and +0.70, +0.43 and +0.20 !lA, respectively. The maximum (-0.90 and +0.70 !lA) and the minimum cunents (-0.14 and +20 !lA) occurred broadly during the afternoon and early morning hrs (LT), respectively. It is, thus, clear that the PDC activity during thunderstorms at Pune exhibits a typical diurnal cycle with maximum during the afternoon and minimum during the early morning hours.

The other results of this study is concerned with the charges of both signs (±mC) reaching the earth by PDC and their diurnal variation. Figures 1 (c) and l(d) show the diurnal variation of storm averaged values of positive and negative charges (±mC), respectively. A computerized smooth curve is drawn through each set of these data [Fig. 1{(a)-(d)}]. It is noted that the trend of variation of curve through currents and charges of one polarity shows anti-phase relationship with that of the opposite polarity. This observation is interesting and suggests that although the diurnal maximum of negative and positivi"

activity, by and large, appears in the afternoon_ i:nere is a strong mix of their activity¹⁶ that is well-separated by time. The present results [Fig. 1 {(c) and (d)}]

show that the storm averaged diurnai total negative (:LmC) and positive (l:mC+) charges recei·ved by the earth's surface were -14.5 mC and + 11.5 nzC, respectively. The storm averaged diurnal maximum, mean and minimum of these charges were -0.54. -0.27 (std = 0.13) and -0.04 mC; and +0.61, +0.24 (std = 0.12) and +0.07 mC, respectively. This shows that the net charge [l:mC - mC+] received by the earth was negative although the positive charges were also present. It may be inferred from these results that the diurnal features of cunents and charges are closely associated with the time of activity of the thunderstorms over this part of the Indian sub- continent.

224 INDIAN J RADIO & SPACE PHYS, AUGUST 2003

1.0

(!J 0.9 (a)

z a: ^<(

::J "-0.8

0 _j - +ve current (~A)

ii ~ 0.7 0 a:

t;; ~ 0.6 a: ~

~ :Q 0.5

U::J

~ ~ 0.4

" a:

~ ~ 0.3 w a:

~ ~ 0.2 a:o ~ 0.1

<(

0.0

0 18

1.0

"

^{z 0.9 (c)}

ii' :ou 0.8 o E

::;; .

a:"' 0.7 0 - ' .... <

(/)it _0.6 a: w

- +ve charge (mC)

w ....

Q. z w[( 0.5 '-':::>

a:o 0.4

~J: ua:

"> w .... 0.3 + a:

w <

"::;) _{< 0} ^0.2

a: w

> 0.1

<

0.0

0

In LST

in LST

24

1.0. - - - -- - - - - -- - - ,

(!J z 0.9 Ci:<

::J "

0 _j 0.8

::: <(

~ ~ 0.7 t;~ a: ~ 0.6 wen ~ ~ 0.5 cO ~ ~ 0.4

> w

~ ~ 0.3

(!J <(

<(::J

ffi ^{0 0.2}

>

<( 0.1

0.0

"

^z _ii' ^0.9

::JC.)

o E _0.8

:::; .

a:"'

~ ci ^0.7

(/)it a: w w .... 0.6

Q. z w;;:

""'

0.5 a:o ~J:

u a: 0.4

g! ~

•II: 0.3 w <

""'

<0 a: 0.2 w >

< 0.1 0.0

- -ve current (~A

(d)

- -ve charge (mC)

0

1n LST

m LST

TIME,hrs LT

Fig. 1- Storm averaged local diurnal variation of PDC of (a) positive and (b) negative polarity; and storm averaged diurnal variation of charge of (c) positive and (d) negative polarity, respectively

The present results [Fig. 1 (a)-(d)] are compared with the results obtained on the basis of several years of observations at Kew^17.19'30 at Nigeria and also with those at Nagycenk, Hungary¹⁹ . As displayed in their study Whipple and Scrase¹⁷ showed a clear predominance of negative charge flow in the PDC to the earth, especially during the afternoon hours. At Nagycenk¹⁹ also, a maximum of the hourly averaged negative and positive charges appears in the afternoon with a minimum in the early morning. In all these cases the PDC and charge maximum and minimum occur during the local afternoon and in the early morning hours, respectively. We have noted in this study that, by and large, the variations of current and charge received by the earth are comparable with the studies mentioned above. It may, therefore, be inferred that although climatic and meteorological conditions and the tropical location of Pune differ

from those of the earlier studies, the basic nature of local diurnal variation of PDC is similar everywhere.

In view of the present shortcomings of the data. it may look amiss to present the values of the diurnal mean amplitude ratios for currents and charges. The diurnal mean amplitude ratio for currents worked out to be 1.4 and for charges 2.1. These ratios are much larger than those of Carnegie (0.35) and for the predicted global circuit response for PDC (0.47), and are quite larger than those for the predicted cloud-to- ground lightning (1.0) obtained elsewhere^16. But in the present case, it is also noted that the data for foul weather conditions (other than thunderstorms) were not available. Possibly, the present result of larger values of mean amplitude ratios would have been then more closer to those studies^16, suggesting somewhat better agreement in phase comparison with those studies and that, at Pune, diurnal variation of PDC is

strong. The present result suggests that the point discharge current source is universally uniform in local time over the diurnal cycle, and necessarily forms a phase of charge exchange during a thunderstorm.

4.2 Diurnal variation of the ratios of negative to positive charges deposited to the earth's surface by PDC

The diurnal variation of the ratios (R

=

mCJmC+) of the negative to positive charges reaching the earth's surface by PDC at Pune has been examined.

The diurnal ratios have been computed for the 41 concurrent QHI values of storm averaged negative and positive charges appearing during 1330-2400 hrs LT of the diurnal period [see Fig. 1(c) and (d)]. The present analysis has shown that R maximum, mean and minimum were 5.34, 1.38 (standard deviation

=

0.87) and 0.27, respectively. The median of these 41 ratios was 1.13. Some details of these 41 ratios are more revealing. It is noted that out of the total 41 ratios, 22 ratios were larger than 1.0, with their mean

=

2.04, and 19 ratios were smaller than 1.0 with their mean

=

0.66. The diurnal time associated with these two sets of ratios (R > 1.0, and R < 1.0) indicated that larger values of R (> 1.0) occurred mainly during 1330-2100 hrs LT, and smaller values of R (<1.0) occurred mostly beyond 2100 hrs LT. In some other studies^18•19 the sustained occurrence of positive PDC during different hours of the local time was also reported. Thus, the present analysis presumably suggests that the diurnal surplus of negative and positive PDC may be a characteristic phenomenon of the local thunderstorm activity at this place.

4.3 Examination of prevalence of positive electrical activity by late night hours

In order to study whether the active period of PDC of one polarity has a preferred diurnal time of occurrence over the other, we have drawn a computerized polynomial smooth curve of 7 degree in each block of Fig. 1 through the data points covering the major period of its activity, mentioned earlier. It is noted that the curve in Fig. 1(a) is similar to curve in Fig. 1(c). Similar is the nature of the curves of Fig.

1 [(b) and (d)]. A comparison between the two sets of curves [Fig. 1{(a), (c)} and Fig. 1{(b), (d)}] is interesting and revealing. It is noted that the maximum and minimum in one set shows anti-phase relationship with that in the other. A close examination of the trends of smooth curve [Fig.

l{(a),(c)}; Fig. 1{(b),(d)}] beyond 2100 hrs LT, QH84, indicates sustained higher values of positive

PDC and charges than those of the negative ones.

This observation suggests that, by and large, the active period of PDC of one polarity has a preferred time of occurrence over the other, and also that by late night hours the posrtrve electrical activity predominates over the negative ones.

This point of view was further examined in details by segregating the diurnal period into three time sections, i.e. 1400-2100 hrs LT; 2100-0800 hrs LT;

and 0800-1400 hrs LT. These time sections broadly represent the storm activity into afternoon (AN), night and day time, respectively. Tables 2 and 3 present the maximum, mean and minimum values of PDC and charge of both polarities, respectively, during the three time sections. It is noted from Tables 2 and 3 that the positive mean and minimum values of current and charge are observed to be higher than those of the negative ones only during the 2100-0800 hrs LT interval and not at other time. This observation points out the sustenance of positive electrical activity by late night hours. Although our understanding of the cloud electrification, in general, and the excess of positive electrical activity, in particular, is limited, reports of their increased occurrence are available elsewhere³¹. Further studies are needed for clear understanding of this behaviour.

4.4 Seasonal variation of PDC and air temperature at Pune The monthly mean amplitude of PDC during April-October at Pune is compared with the monthly mean maximum surface air temperature. Previous studies^23•32 have obtained positive correlation between tropical and globally averaged air temperature and

Table 2-Details of the variation of PDC of both polarities during different time intervals

Polarities during

PDC AN Night Morning

j.I.A (1400-2100 (2100-0800 (0800-1400 hrs LT) hrs LT) hrs L T) Maximum -0.901+0.70 -0.70/+0.65 -0.90/+0.57 Mean -0.56/+0.47 -0.33/+0.39 -0.53/+0.36 Minimum -0.29/+0.23 -0.14/+0.20 -0.33/+0.23

Table 3-Details of the variation of charge of both polarities during different time intervals

Polarities during

Charge AN Night Morning

mC (1400-2100 (2100-0800 (0800-1400 hrs LT) hrs LT) hrs LT) Maximum -0.54/+0.61 -0.53/+0.48 -0.57 1+0.24 Mean -0.33/+0.26 -0.19/+0.23 -0.27 1+0. I 6 Minimum -0.14/+0.08 -0.041+0.09 -0.14/+0.07

226 INDIAN J RADIO & SPACE PHYS, AUGUST 2003

global electric circuit, both on semi-annual and annual time-scales. Since PDC is a dominant parameter of the global electric circuit, it is believed that a comparison between seasonal air temperature and PDC is possible and worthwhile. The monthly data of PDC for the thunderstorms during April-October, referred earlier, are used to obtain the storm-averaged monthly mean amplitude of PDC, while the monthly mean maximum surface air temperature data for Pune station are taken from the climatological normals

maximum temperature and cloud electrification, and hence PDC.

It is realized that in the present study the data from one station, and the length of the data and its demarcation into premonsoon-monsoon seasons pose a constraint on the evaluation of the results. However, it is believed that the present analysis strongly supports a connection between temperature and PDC, a vital element of global electric circuit.

published by the India Meteorological Department 4.5 Examination of association between duration and

(IMD), Pune. amplitude variation ofPDC

The monthly variation of the two parameters is The per-storm current duration and average shown in Fig. 2. It is noted that the PDC curve tracks amplitude variation of PDC are very much dependent quite closely with that of the temperature. The on the distance between tl}e station and the electrical correlation coefficient between the two parameters is charge centres of a thunderstorm cloud. Storms that 0.80. The trend of the seasonal variation of the two pass directly overhead or are closer to the station will parameters (Fig. 2) is compared with the earlier have longer duration and stronger cutTent amplitude studies²³. It is observed that although we do not have than when the centres are a few kilometres away. It is data for March, there is a high degree of similarity in proposed to examine the above point -of view using the seasonal variation in both the sets. the durations and current amplitudes data from the

Figure 2 shows that the temperature as well as the present set of observations.

amplitude of PDC attains its minimum seasonal value For each storm considered in this study, the in July and August, which are the mid-monsoon duration of total positive and negative current and months when the monsoon activity is usually at its their storm averaged amplitude were computed. Class peak. Under such conditions the atmosphere is moist intervals of 10-min durations (1-10, 11-20, 21-30, neutral and the cloud electrification is much less ... ) of PDC were formed. The storm-wise average pronounced than in the premonsoon season when it is amplitude of current of each polarity was put in conditionally unstable and electrically more active²³. appropriate class interval according to its duration. The result of the above analysis suggests that there is This practice was followed for all the storms, and at a strong seasonal association between surface air the end, the mean amplitude of the current for each

0.8.--- ---,42

<(

""

c5 0 0..

z <(

w :;;

>

..J :I:

...

z 0 :;;

0 w

~ a:

0.7

0.6

0.5

0.4

~ 0.3

<(

:; a:

~ 0.2

"'

•

~ \

Apr May June July MONTHS

•

Aug Sep Oct

0

0

w 40 g;

...

_<(

a: w :;; w

38 ...

w ~

IL a:

36 ~

:;;

:::l :;;

34 ~ ::;;

z <(

w :;;

>

32 :i!

...

_z

0 ::;;

Fig. 2-Seasonal variation of monthly mean amplitude of PDC and monthly mean maximum surface air temperature during April-October at Pune

1.4

< _::1. ^Negative

':i 1.2 Y =A+ B1*X + B2*X"2 + B3*X"3

Co7•

a: 0 1-(/)

a: ^1.0

11. w

•

()

. /

c 0.8

11.

u.

0 / Positive

w 0.6

c _{• -- •- • Current}

::I

·- .

1-

. ^/

^.~

:::i 11. 0.4-

~ ---·---·---· ---·---· ---

:: <

1

w 0.2-

(!) <

a: w

> _0.0

<

~

^{I I I I I I I}

0 20 30 40 50 60 70 80 90 100 110

PDC DURATION PER STORM, min

Fig. 3 -Association between duration and amplitude variation of PDC during thunderstorms at Pune [R-square(COD)=0.55174; SD=0.07108; N=10; ?=0.1602

Parameter A Bl B2 B3

Value 0.06423 0.01954 -3.4042E-4

1.84713E-6

Error 0.13794 0.01034 2.13268E-4 1.27887E-6

class interval of duration was worked out. Figure 3 shows the variation of the mean current of both polarities with the PDC duration for each class interval of time. A computerized best-fit curve of third degree has been drawn through each set of data points in Fig. 3. The details of the curve fit are also furnished in the figure. It is noticed from these curves that the trend of variation of current of both polarities is quite similar. But the amplitude of the negative PDC is systematically larger than that of the positive one for each class interval of duration. For current duration in excess of 60 min the negative amplitudes of PDC are noted to be larger than those of the positive one by 1.4 times to 2.0. We are reminded of the curve by Malan³³ showing the variation of the electric field at the ground surface due to considerations of the model cloud charge centres in a thunderstorm cloud³³. The curve shows the result of such a calculation^33, and the present ones for the positive and negative currents show a high degree of similarity. This similarity suggests that the distance of the thundercloud corresponding to 100-120 min duration of PDC (Fig. 3) could be within - 4 km from the observation point and the portion of the curve that is nearly parallel to the x-axis could be taken as 8-12 km. For distances greater than 12 km the current is

either zero or not detectable. This observation appears to be consistent with the usually reported detection range (- 12 km) for the individual corona²²•25

5 Conclusions

It is believed that a systematic investigation of the

above aspects of PDC variation may yield useful information on the locations of the active charge centres of a storm cloud during different seasons at different times over a locality, if the storm-wise PDC data are grouped on the basis of seasons, their diurnal time of occurrence, their active or dissipation stages, etc. It is proposed to conduct this study as an extension of this work in future.

Acknowledgements

The authors are thankful to the Director, IITM and Head, P M & A Division, IITM, Pune, for their constant supports in doing the research activity.

References

1 Worrnell T W, Proc R Soc London SerA, (UK), 127 (1930) 567.

2 Worrnell T W, Q J R Meteorol Soc (UK), 79 (1953) 3.

3 Kirkman J R & Chalmers J A, J Atmos & Terr Phys (UK), 10 ( 1957) 258.

228 INDIAN 1 RADIO & SPACE PHYS, AUGUST 2003

4 Sivaramkrishnan M V, Indian J Meteorol & Geophys, 8 (1957) 379.

5 Kamra A K & Varshneya N C, J Geomagn & Geoelectr (Japan), 20 ( 1968) 221.

6 Kamra A K, J Atmos & Terr Phys (UK), 31 (1969) 1281.

7 Kamra A K, Geophys Res Lett (USA), 16 (1989) 127.

8 Ette A I I & Utah E U, J Atmos & Terr Phys (UK), 35 (1973) 785.

9 Ette A I I & Utah E U, J Atmos & Terr Phys (UK), 35 (1973) 1799.

10 Rao AM & Ramanadham R, PAGEOPH (Switzerland), 117 (1979) 904.

11 Standler R B & Winn W P, Q J R Meteorol Soc (UK), 105 (I 979) 285.

12 Vonnnegut B, J Geophys Res (USA), 89 (1984) 1468.

13 Asuma Y, Kikuchi K, Tanaguchi T & Fuji S, J Meteorol Soc Jpn (Japan), 66 (1988) 473.

14 Engholm CD, Williams E R & Dole R M, Mon Weather Rev (USA), 118 (1990) 470.

15 Chalmers J A, Atmospheric Electricity (Pergamon Press, Oxford), 1967.

16 Williams E R & Heckman S 1, J Geophys Res (USA), 98 (1993) 5221.

17 Whipple F 1 W & Scrase F J, Geophys Mem Lon (UK), 68 (I 936) 1.

18 Ette A I I & Oladiran E 0, PAGEOPH (Switzerland), 118 (1980) 753.

19 Marez F & Bencze P, J Atmos & Solar Terr Phys (UK), 60

(1998) 1435.

20 Markson R, Nature (UK), 320 (1986) 588.

21 Rutledge S A, Williams E R & Keenan T D. Bull A111 Meteorol Soc (USA), 73 (1992) 3.

22 Williams E R, Rutledge S A, Geotis S G, Renno N.

Rasmussen E & Rickenbach T, J Atmos Sci (USA), 49 ( 1992) 1386.

23 Williams E R, Mon Weather Rev (USA), 122 ( 1994) 1917.

24 Manohar G K, Kandalgaonkar S S & Sholapurkar S M, Curr Sci (India), 99 (1990) 367.

25 Manohar G K, Kandalgaonkar S S & Sholapurkar S M. Mon WeatherRev(USA),119(1991)3104.

26 Manohar G K, Kandalgaonkar S S & Tinmaker M I R, J Geophys Res (USA), 104 (1999) 4169.

27 Kandalgaonkar S S, Tinmaker M I R & M K Kulkarni. J Atmos Electricity (Japan), 21 (2001) 95.

28 Rao A M & Raman P K, Indian J Meteorol & Geophvs. 12 (1961) 103.

29 Raman R K & Raghavan K, Indian J Meteorol & Geophvs.

12 (1961) 115.

30 Israel H, Atmospheric Electricity (Israel Program for Scientific Translations, Jerusalem), 1973, p. 796.

31 Williams E R,J Atmos & Solar Terr Phys (UK). 60 (1998) 689.

32 Williams E R, Global circuit response of temperature on distinct time scales, in Status Report, Parsons Lab.

Massachusetts Institute of Technology, Cambridge.

Massachusetts, June 1997, pp. 1-13.

33 Malan D 1, Physics of Lightning (Th~ English University Press, London), 1963, p. 46.