Studies on the seasonal variation of atmospheric electricityparameters at a tropical station in Kolkata, India

S.S. De a,n, Suman Paul a, S. Barui a, Pinaki Pal b, B. Bandyopadhyay a, D. Kala a, A. Ghosh a

a S.K. Mitra Centre for Research in Space Environment, Centre of Advanced Study in Radio Physics and Electronics,University of Calcutta, Kolkata 700009, Indiab Department of Physics, Maharaja Bir Bikram College, Agartala, Tripura (W) 799004, India

a r t i c l e i n f o

Article history:Received 24 March 2013Received in revised form9 September 2013Accepted 11 September 2013Available online 18 September 2013

Keywords:Global electric circuitFair-weather atmospheric electricityVertical potential gradientAtmospheric conductivity

a b s t r a c t

The paper deals with the analyses of the atmospheric vertical potential gradient (PG) from the ground for90 fair weather days during 2006–2009 measured at Kolkata (Lat: 22.561N). The variations of PG havebeen studied extensively to investigate their values during monsoon and winter seasons. Higher values ofPG at Kolkata are observed due to higher abundance of pollutant particles. The observed PG arecompared with the results of Potsdam station (Lat: 521N) and Johannesburg station (Lat: 261S), with9 years data and 2 years data respectively. The correlations studies are carried out among PG, PDC (PointDischarge Current) as well as negative and positive carrier conductivities. The corresponding correlationcoefficients are obtained as 0.93, �0.842 and �0.844.

& 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Thunderclouds in the troposphere updraft the charges produ-cing current into the ionosphere and magnetosphere that main-tains an electric potential difference of nearly 250 kV between theEarth and the ionosphere which produces vertical electric poten-tial gradient (PG) on the surface of the Earth. At the ground, this ismaintained by the negative quasi-steady charge on the Earth andby the ‘electrode effect’. It is due to the buildup of positive chargein the air near the surface of the Earth and this is because of theflow of positive charge from the ionosphere to negatively chargedEarth (Marshall et al., 1999). Negative charges do not flow from theEarth because of the want of any mechanism except the process ofradioactive phenomena that yield the emanation from the litho-spheric origin. Vertical electric field of about 100–120 V m�1 isfound at the ground surface of the Earth. Among different sourcesof electromotive force driving the global circuit, thunderstorms areconsidered to be the most powerful ones. Dynamo-interactionbetween the solar wind and the magnetosphere, and the dynamoeffects of atmospheric tides in the thermosphere are considered assources not as powerful as the thunderstorm generated by thecreation of electric field driving the global circuit at the loweratmosphere. Further, the Earth–ionosphere potential differencedirects the air–earth current downward from the lower region of

the ionosphere to the ground surface which varies in accordancewith the ionospheric potential and columnar resistance.

Major thunderstorm activities of the globe and local environ-mental factors maintain electric PG at any point on the Earth'ssurface. Various models for thundercloud electric field have beenpresented to investigate their behaviour in the region between thesurface of the Earth and the lowest layer of the ionosphere(Wilson, 1920; Park and Dejnakarintra, 1973; Yeboah-Amankwah,1989; Kumar et al., 1995; Tonev and Velinov, 1996).

It is worth-mentioning that pollutant particles due to smokefrom combustion processes (domestic and industrial origin), reac-tion of natural and anthropogenic gaseous species, and wind-blown dust can reduce the conductivity giving higher value ofPG (Prospero, 1984; Harrison and Aplin, 2002, 2003). Aerosolparticles, which are smaller than 5 μm in diameter, tend to formstable suspension in air (Malm et al., 1994). The nuclei present(pollutant particles: artificial or natural) combine with the ionsand decrease the concentration or immobilize the small ions,thereby reducing the conductivity. The impact of fossil fuels (CO2

emission) within the pollutants is also responsible for higher trendin the value of electric field than the fair-weather value.

There are changes in ionization rate, recombination and attach-ment rate, and also different meteorological conditions, whichaffect the variation of electrical conductivity. As a result, atmo-spheric electric field and air–earth current are affected. The inter-dependence of these parameters is important. Meteor showers,solar flares, sudden meteorological disturbances like severe cyclo-nic storms can affect the parameters of global electric circuits

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/jastp

Journal of Atmospheric and Solar-Terrestrial Physics

1364-6826/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.jastp.2013.09.006

n Corresponding author. Tel.: þ91 33 2350 5829; fax: þ91 33 2351 5828.E-mail address: de_syam_sundar@yahoo.co.in (S.S. De).

Journal of Atmospheric and Solar-Terrestrial Physics 105-106 (2013) 135–141

through the variation in atmospheric conductivity (Herman andGoldberg, 1978). The variations in the meteorological conditionscan give rise to rapid changes in the value of PG during rain,showers, snow as well as fog. The changes in wind direction canalso produce variations in PG (Bennett and Harrison, 2007). Rainand rain-clouds are often highly charged. As a result, the value ofPG may increase. The absolute value of PG and its variation arelargest during thunderstorms. Moreover, surface PG should alsodepend on change of pressure, temperature and formation ofdense fog which governs the conductivity of the medium. Thus,the meteorological effects play some role in the enhancement ofPG (Anbar, 2006; Founda et al., 2007; Kachakhidze et al., 2009).Hence, seasonal variations are to be taken into consideration.

In many observations, the diurnal variation of PG matchesclosely with the Carnegie curve, which is the generally acceptedglobal unitary variation of PG on Earth with a maximum around19:00 h UT (Universal Time) and a minimum around 04:00 h UT(Takagi and Kanada, 1972; Kasemir, 1972; Yeboah-Amankwah,1989). But, based on the observations of the atmospheric electricfield in the Indian Ocean, Bay of Bengal and Arabian Sea, Kamraet al. (1994) showed large changes in PG value, in which 40-dayaverage field curve showed a maximum at 10:00 h UT and aminimum at 00:00 h UT with a small secondary peak at 19:00 hUT. The observations at the Indian station Maitri at Antarctica (Lat:70.751S) are also in contrast to the unitary diurnal variation ofCarnegie curve (Deshpande and Kamra, 2001). The diurnal andseasonal variations of ground surface electric field over Pune (Lat:18.321N) during fair-weather days of 1993 are measured and theinfluences of meteorological factors over the observed results havealso been reported (Latha, 2003). Some recent measurements ofelectric field for 69 fair weather days during local summer at theIndian station at Antarctica, Maitri, showed the maximum andminimum period of occurrence similar to the Carnegie curvealthough the field values are much higher than the Carnegieresults (Panneerselvam et al., 2007).

Analysis of daily and seasonal variations at different locations onthe globe is important as it will provide additional information for thestudy of the GEC. Our attempt in the present paper is to study thevariations of PG in fair-weather conditions at the near-equatorialstation Kolkata. The seasonal variations in PG have extensively beenstudied to investigate mainly the difference between monsoon (whenthe Sun is in Northern Solstice) and winter (when the Sun is inSouthern Solstice) values. The results are compared with thosereported by others at various stations. The observed records of pointdischarge current (PDC), conductivity and PG during the period fromJanuary 2006 to December, 2009 at Kolkata are analyzed. Thecorrelations between PG and PDC and between PG and near Earthatmospheric conductivity have also been studied.

2. Experimental arrangement

The vertical electric field has been measured on a continuousbasis using an ac field-mill placed at a height 26 m above theground. Its aluminium rotor plate is 12 cm in diameter while thestator has the same dimension. The alternating signal fromthe field-mill is being amplified using a signal processor havingone-second time constant. IC LF356N has been used at the inputstage of the amplifier because of its high input resistance (�1012 Ω)and good signal to noise ratio. The r.m.s. value of the amplified signalis used to find the required electric field from the calibration chart. Ithas been calibrated in a vertical field set-up between two largealuminium cover plates, electrically isolated at a given potential,through a fixed distance between them. The outer shield of the field-mill is grounded properly to ensure protection from possible fielddistortions. The sensitivity of the field-mill is measured to be

(0.3370.03) V m-1. The output has been recorded by a digital dataacquisition system at a sample rate of one data per second.

The continuous measurement of PDC has been made by using asteel wire having diameter 3 mm and length 8 cm, one end ofwhich is tapered to a sharp edge. The tip of the sharp edge is about0.02 mm. The other end is soldered to a co-axial cable which ismade perfectly insulated by Teflon. The entire junction is tightlycovered by a Teflon insulated wire. The cable is surrounded bythermoplastic polystyrene which is coated by a very thin honey-comb winding copper wire. The external surface is kept within thecylindrical plastic cover. This process ensures efficient heat insula-tion also. The other end of the cable is connected to the receivingsystem. The pointer is erected on a wooden support which is at aheight of 8 m from the ground. Proper precautions are taken toavoid any other objects with sharp edges at the site. The transientresponses from the tip are amplified. The overall gain of theamplifier is around 40 dB. IC LF356N is used here to ensure goodsignal to noise ratio. The output of the amplifier is recorded at asample rate of 1 data per sec, through a data acquisition systemthat uses a PCI 1050, 16 channels 12 bit DAS Card (Dynalog), having12 bit A/D converter, 16 digital inputs and outputs. One of theinput–output channels is used for PG measurement and another isfor PDC signal measurement. The recorded data are analyzedthrough Origin 5.0 software. A set of 15 data have been averagedand then plotted. A block diagram is shown in Fig. 1.

The conductivity is measured near ground through a Gerdiencondenser, where air between the electrodes (two co-axial cylinders)is supplied by a fan. Air-ions with desired polarity and mobility areforced by electric field to supply their charge to the collecting electrodewhich gives the generating current I¼neQ. Here, Q is the amount of airflow through the electrode (�0.0022 to 0.0026) m3 s�1 and e is1.609�10�19 C. Current is measured by an electrometer (Kolarz et al.,2012). It is represented as the concentration of air-ions cm-3.

Polarizing voltage (U) and air flow determine the criticalmobility mc of the measured ions given by the following expres-sion:

μc ¼VsðR2

1�R22ÞlnðR2=R1Þ2LU

where, R1, R2 are the radii of the polarizing and collectingelectrodes; L is the electrode length and Vs is air-flow speed. Theinstrument is used for air conductivity measurements undervarying concentrations and scanning of air-ions through mobility.Alternating ion polarities would require alternative supply ofapplied polarizing voltage that produces capacitive current spikes(Kolarz et al., 2012; Vojtek et al., 2006).

3. Observational results and discussions

The PG shows a diurnal variation with maxima and minima.Values are found to be higher than the usual trend of fair-weatherresults. Fig. 2 depicts diurnal variations averaged over 90 fair-weather

Fig. 1. Block diagram of receiver system.

S.S. De et al. / Journal of Atmospheric and Solar-Terrestrial Physics 105-106 (2013) 135–141136

days available during the period from January 2006 to February 2009.Standard deviations are shown by error bars. The measured diurnalvariation in PG was obtained by 15min average. Maximum PGexhibits 193 Vm�1 that occurs at about 04:00 h UT (Universal Time)(09:30 h Local time (LT), late morning period). It is followed by asecondary maximum between 08:00 and 08:30 h UT. It thendecreases showing a minimum at around 10:30–11:00 h UT (16:00h LT) which is before the local afternoon period. Another secondarymaximum occurs at about 16:00 h UT. The potential level starts todecrease gradually and exhibits lowest value at around 00:00 h UT(05:30 h LT), about half an hour before sunrise. It then rises to amaximum value with a repetition at around 04:00h UT following thenew cycle. So the fair day features consist of one principal maximumand two secondary maxima.

Our observation shows a marked variation from Carnegie'soceanic field curve, which is the generally accepted global unitaryvariation of electric field on Earth. Maxima and minima inCarnegie curve differ from the present results of 90 days averagefield curve. The overall average, irrespective of seasons, shows aprimary maximum at about 04:00 h UT (09:30 h LT) and onesecondary maximum at about 16:00 h UT (21:30 h LT) instead of asingle maximum at 18:00–19:00 h UT of Cranegie curve. Similardeviations have been reported by Kamra et al. (1994) andDeshpande and Kamra (2001). Sunrise and its prolonged effectsare responsible for the primary maximum at about 04:00 h UT(09:30 h LT). Sunrise effect is explained by a two layer electrodemodel (Hoppel et al., 1986; Marshall et al., 1999). By sunrise, theupliftment of dense electrode layer due to convective mixingabove the sensing height of 2 m increases the value of PG(Kamra, 1982). Near Earth surface conductivity, seasonal variationof ionospheric potential and distance from active thunderstormareas may also influence the vertical PG (Takagi, 1977). Thecommonly accepted thunderstorm occurrence frequency curve ofWhipple and Scarce (1936) with the maximum at around 19:00 hUT and minimum around 04:00 h UT is the average of three majorcontinental thunderstorm activity centres viz., Asia–Australia,Africa–Europe and America (Orville and Henderson, 1986). More-over, each continental average was taken over 81 years.

Our results also indicate a tendency to manifest the thunderstormactivity over the Asia–Australia region which becomes maximum at08:00 h UT. It is worthwhile to mention that the measurements at theAsiatic tropical zones exhibit the resemblance of peak thunderstormactivity at 08:00 h UT of Asia–Australia region. Thus, in a tropicalregion, continental thunderstorm activity plays an important role inmodulating the global electric circuit. Observations of global lightningdistribution during January 1998–February 2009 taken from LIS

Fig. 2. Hourly average diurnal variation of Potential Gradient (PG) of 90 fairweather days at Kolkata during January 2006–February 2009. Error bars representstandard deviation.

Fig. 3. Vertical electric Potential Gradient (PG) of the atmosphere at Kolkata for theperiod June 2008-January 2009 is shown by blue coloured curve, while the blackcoloured curve depicts the results for the period from June 2005-January 2006.Corresponding standard deviations are shown by error bars. (For interpretation ofthe references to colour in this figure legend, the reader is referred to the webversion of this article.)

Fig. 4. Diurnal variations of Potential Gradient (PG) at different seasons during January 2006–February 2009. Standard deviations are represented by error bars.

S.S. De et al. / Journal of Atmospheric and Solar-Terrestrial Physics 105-106 (2013) 135–141 137

(Lightning Imaging Sensor), NASA, confirm that the thunderstormactivity in the Himalayan region is more prominent than in the Asia–Australia region (http://thunder.nsstc.nasa.gov). The Asia–Australia haslong been thought of as one of the largest thunderstorm producingregions, but recent records confirm that the thunderstorm from theHimalayan region is also strong enough to be one of the distinguishedlightning centres all over the world. So the peak at 04:00 h UT ishigher than that at 08:00 h UT.

After morning period, solar heating near the surface of theEarth initiates convective instability which returns the PG value toits typical fair-weather value (minimum around 10:00–12:00 hUT). But the secondary maxima around 16:00 h UT (21:30 h LT) isneither due to GEC nor sunrise effect. It may be due to regionalthunderstorm activity mainly in the Himalayan region.

Fig. 3 shows the comparative study of vertical electric potentialgradient of the atmosphere over Kolkata for the period June 2008to January 2009 (blue coloured curve) along with the results ofearlier data recorded at this centre for the same period of duration,June 2005 to January 2006 (black coloured curve). Standarddeviations from the mean are represented by corresponding errorbars. The two periods of study are separated by a gap of 3 years.The figure depicts that the nature of diurnal variation remainsalmost the same for the two periods. But with the progress intime, the PG shows a tendency to decrease on an average.

The decrease of both kinds of ions at the ground level is one of thesignificant reasons for the higher value of PG at Kolkata. The atmo-spheric turbulence forces the positive space charges to higheraltitudes against the forces of the existing electric field. It can betaken as the generator effect of the turbulence which affects theatmospheric electric field locally. The decrease of ions in the process oftransportation lowers the conductivity. Kolkata is a densely populatedcity surrounded by small and large scale industries. Air is greatlyinvaded by pollutant particles emitted from various industries. Fineparticles having a diameter less than 1 mm (Aitken nuclei) aredistributed in air. They can capture both kinds of ions. Since ionsare attached to pollutant particles of comparatively large mass, themobility of the ions decreases (Pawar and Kamra, 2000; Tinsley,2000). So near the Earth, conductivity is low. The magnitude ofrelative abundance of different aerosol particles as well as thepresence of Aitken nuclei distributed in air at this tropical station islower than its value during the period 5–6 years earlier (Raj et al.,1997). The existence of their variation introduces a change in thevalue of vertical PG in the atmosphere. The variation of energyconsumption of traffic in Kolkata would contribute to a high Aitkencount resulting in changes in atmospheric dispersion that also reducethe conductivity of the medium. During the period June 2008–January2009 in comparison to the period of duration, June 2005–January2006 many industries surrounding Kolkata have been shifted to thesuburban areas and some were abandoned. The reduction in thevariation of energy consumption of traffic (oil and gasoline) and fossilfuels (CO2 emission) during the later period of study increases theconductivity of the medium near the surface of the Earth. Hence thevalue of PG shows a tendency to decrease on an average.

In order to explore the seasonal effect of diurnal variations of PG,the whole year is divided into four seasons, viz., winter (December,January and February), pre-monsoon (March, April and May), mon-soon (June, July and August) and post-monsoon (September, Octoberand November). Fig. 4 shows diurnal variations of PG at differentseasons averaged over the period from January 2006 to February2009 along with their standard deviations plotted as error bars.During all the seasons, the first maximum which is very muchprominent occurs at around 03:30–04:00 h UT (local late morninghour). During winter and pre-monsoon, the first secondary maximumobtained around 08:00–09:00 h UT is comparable to the principalmaximum. The second secondary maximum occurs at 13:30 h UT,19:00 h UT, 14:30 h UT and 16:00 h UT respectively during winter,

pre-monsoon, monsoon and post-monsoon seasons. The values at theprincipal peak of PG in winter, pre-monsoon, monsoon and post-monsoon are 195, 190, 170 and 180 Vm�1, respectively. It is evidentthat the primary maximum PG value during monsoon is 170 Vm�1

which is a comparatively low value during the year. It is due to thefact that during the monsoon season aerosol concentration in thenear Earth surface medium is reduced. Also, the patterns in Fig. 4

Fig. 5. Mean value of vertical electric Potential Gradient (PG) against the number ofdays observed during 2006–2009.

Fig. 6. Diurnal variation of Point Discharge Current (PDC) and Potential Gradient(PG) at Kolkata during monsoon and winter of 2006, 2007, 2008 and 2009.

S.S. De et al. / Journal of Atmospheric and Solar-Terrestrial Physics 105-106 (2013) 135–141138

indicate that the sunrise effect affects all the seasons almost similarly.Although, in Kolkata, which is a site of observation in the tropicalregion of Northern hemisphere, significant variations in local sunrisetime occur during different seasons, no shift in principal maxima ofPG value is found. One such shift in the morning peak about 45 minwas reported from Suva, Fiji station in Southern hemisphere duringWet and Dry seasonal measurement (Kumar et al., 2009).

Fig. 5 is plot of mean value of vertical electric potential gradientversus the number of fair-weather days observed during the

period of study (2006–2009). The potential gradient is higharound 152 V m�1 which is also the mean value at Kolkata.

The hourly variations of PDC and PG at Kolkata averaged overmonsoon season and winter are shown in Fig. 6 separately for theyears 2006, 2007, 2008 and 2009. The PG and PDC show the sametrend in their diurnal variations. During monsoon, minimumand maximum values of PDC are 2.0�10�10 A m�2 and 2.55�10�10 A m�2. During winter, the PDC varies between a minimumvalue of 2.2�10�10 A m�2 and a maximum value of 2.62�10�10 A m�2. The minimum value is obtained at 24:00 h UT (halfan hour before the sunrise) and maximum value is obtained around04:00 h UT (local late morning hour). The PG is observed to vary from130 Vm�1 to 170 Vm�1 during monsoon and from 170 Vm�1 to195 Vm�1 in winter. The correlation coefficient between PDC and PGis remarkably high as shown in Table 1. The measurements of PG andPDC in this work are two different ways of measuring the same strongelectric field. So, high correlation is expected and data presented inTable 1 also admit the phenomenon. But, the influence of wind speedalways prevails in the process which forces the space charges awayfrom the point of discharge, thereby serving to unshield those thatincrease the coronal current. This effect of the additional parameterenhancing the coronal current if taken into account, the said correla-tion should not maintain one-to-one correspondence as stated(Williams, 2013).

Typical annual variations of PG are shown in the upper panel ofFig. 7 and the negative carrier conductivity (NCC), positive carrierconductivity (PCC) and PDC averaged over the period from 2006 to2009 are shown in the lower panel of Fig. 7. The average range ofvariation of PG over four years is between 160 and 190 V m�1. Thisis associated with the variation of PDC between 2.2�10�10 A m�2

and 2.6�10�10 A m�2. The corresponding variations in NCC andPCC are between 2.2�10�14 S m–2 and 2.5�10�14 S m�2. Thecorrelation coefficient of PG with PDC, NCC and PCC are respec-tively 0.93,–0.842 and–0.844. These are obtained through theanalysis presented in Fig. 8.

Fig. 9 shows the seasonal variation in PG at Kolkata, Potsdam(Northern hemisphere) and Johannesburg (Southern hemisphere)

Table 1Observed correlation coefficient between PDC and PGin monsoon and winter during the period 2006to 2009.

Years Monsoon Winter

2006 0.79 0.842007 0.89 0.912008 0.87 0.882009 0.85 0.87

Fig. 7. Seasonal variations of Potential Gradient (upper panel) and negative,positive conductivities and Point Discharge Current (lower panel) at Kolkata duringJanuary 2006–December 2009.

Fig. 8. Correlations of Negative Carrier Conductivity (NCC), Positive Carrier Conductivity (PCC) and Point Discharge Current (PDC) with Potential Gradient (PG) at Kolkataduring January 2006–December 2009.

S.S. De et al. / Journal of Atmospheric and Solar-Terrestrial Physics 105-106 (2013) 135–141 139

stations. The monthly variation of PG at Kolkata (Lat: 22.561N) isalmost the same as observed at Potsdam (Lat: 521N). It reveals thatthe electric PGs at these two stations yield the maximum valueduring winter and minimum value during the monsoon months.But at the Johannesburg station (Lat: 261S), the picture is just thereverse. It exhibits a maximum during monsoon while a minimumis found to occur during winter (Alderman and Williams, 1996).This is due to the effects of opposite seasonal changes in insolationand convection in the two hemispheres (Williams, 2009).

4. Conclusions

The following conclusions are drawn from the present analyses:

a) The diurnal curve of PG shows a marked deviation fromCarnegei curve.

b) The PG measured at Kolkata is higher than the values observedover the ocean and it is due to the local meteorologicalparameters.

c) With the progress in time, PG showed a tendency to decreaseon an average. It may be due to reduction of pollution inKolkata during the tenure of measurements.

d) Seasonal variation shows that the monsoon level is lower than thewinter level because of a lower amount of aerosol content in air.

e) PG exhibits good correlations with PDC both in monsoon andwinter.

f) The monthly variations in PG observed in the stations ofNorthern hemisphere are identical, whereas, an opposite nat-ure of variations are observed between the results of Northernand Southern hemispheres.

Acknowledgements

The authors acknowledge with thanks the financial supportfrom the Indian Space Research Organization (ISRO) throughS. K. Mitra Centre for Research in Space Environment, Universityof Calcutta, Kolkata, India, for carrying out the study. They are alsothankful to the respected reviewers of this paper for their valu-ables comments and suggestions, the inclusion of which suffi-ciently improved this revised version. Special thanks are due to

Prof E R Williams for some critical discussions related to a queryraised by a reviewer.

References

Anbar, O.M.Y., 2006. Solar eclipse effects on the air layer near the surface overMakkah Region. Department of Meterology, Environment and Arid LandAgriculture, King Abdulaziz University, Jeddah, Saudi Arabia.

Alderman, E.J., Williams, E.R., 1996. Seasonal variation of the global electric circuit.Journal of Geophysical Research 101, 29679–29688, http://dx.doi.org/10.1029/96JD01547.

Bennett, A.J., Harrison, R.G., 2007. Atmospheric electricity in different weatherconditions. Weather 62, 227–283.

Deshpande, C.G., Kamra, A.K., 2001. Diurnal variations of the atmospheric electricfield and conductivity at Maitri, Antarctica. Journal of Geophysical Research106, 14,207–14,218, http://dx.doi.org/10.1029/2000JD900675.

Founda, D., Melas, D., Lykoudis, S., Lisaridis, I., Gerasopoulos, E., Kouvarakis, G.,Petrakis, M., Zerefos, C., 2007. The effect of the total solar eclipse of 29 March2006 on meteorological variables in Greece. Atmospheric Chemistry andPhysics 7, 5543–5553.

Harrison, R.G., Aplin, K.L., 2002. Mid-nineteenth century smoke concentrationsnear London. Atmospheric Environment 36, 4037–4043.

Harrison, R.G., Aplin, K.L., 2003. Nineteenth century Parisian smoke variationsinferred from Eiffel Tower atmospheric electrical observations. AtmosphericEnvironment 37, 5319–5324.

Herman, J.R., Goldberg, R.A., 1978. Initiation of non-tropical thunderstorms by solaractivity. Journal of Atmospheric and Terrestrial Physics 40, 121–134, http://dx.doi.org/10.1016/0021-9169(78)90016-8.

Hoppel, W.A., Anderson, R.V., Willett, J.C., 1986. Atmospheric electricity in theplanetary boundary layer. The Earth's Electrical Environment, 149–165.National Academic Press, Washington, D.C..

Kachakhidze, N., Kachakhidze, M., Kereselidze, Z., Ranishvili, G., 2009. Specificvariations of the atmospheric electric field potential gradient as a possibleprecursor of Caucasus earthquakes. Natural Hazards and Earth System Sciences9, 1221–1226, http://dx.doi.org/10.5194/nhess-9-1221-2009.

Kamra, A.K., 1982. Fair weather space charge distribution in the lowest two metersof atmosphere. Journal of Geophysical Research 37, 4257–4263.

Kamra, A.K., Deshpande, C.G., Gopalakrishnan, V., 1994. Challenge to the assump-tion of the unitary diurnal variation of atmospheric electric field based onobservation in the Indian Ocean, Bay of Bengal and Arabian Sea. Journal ofGeophysical Research 99, 21043–21050, http://dx.doi.org/10.1029/94JD01485.

Kasemir, H.W., 1972. Atmospheric electric measurements in Arctic and Antarctic.Pure and Applied Geophysics 100, 70–74.

Kolarz, P., Miljkovic, B., Curguz, Z., 2012. Air ion counter and mobility spectrometer.Nuclear Instruments and Methods in Physics Research B 279, 219–222.

Kumar, R., Jagdish, R., Singh, V., 1995. Lightning return stroke electric fields aboveground. Journal of Atmospheric and Terrestrial Physics 57, 1247–1254, http://dx.doi.org/10.1016/0021-9169(95)00012-Q.

Kumar, V.V., Ramachandran, V., Buadromo, V., Prakash, J., 2009. Surface fair-weather potential gradient measurements from a small tropical island stationSuva, Fiji. Earth, Planets and Space 61, 747–753.

Latha, R., 2003. Diurnal variation of surface electric field at a tropical station indifferent seasons: a study of plausible influences. Earth, Planets and Space 55,677–685.

Lightning Imaging Sensor (⟨http://thunder.nsstc.nasa.gov⟩).Malm, W.C., Sisler, J.F., Huffman, D., Eldred, R.A., Cahill, T.A., 1994. Spatial and

seasonal trends in particle concentration and optical extinction in the UnitedStates. Journal of Geophysical Research 99, 1347–1370, http://dx.doi.org/10.1029/93JD02916.

Marshall, T.C., Rust, W.D., Stalzenburg, M., Roeder, W.P., Krehbiel, P.R., 1999. A studyof enhanced fair-weather electric fields occurring soon after sunrise. Journal ofGeophysical Research 104, 24455–24469.

Orville, R.E., Henderson, R.W., 1986. Global distribution of mid-night lightning.Monthly Weather Review 114, 2640.

Panneerselvam, C., Selvaraj, C., Jeeva, K., Nair, K.U., Anilkumar, C.P., Gurubaran, S.,2007. Fair weather atmospheric electricity at Antarctica during local summer asobserved from Indian station, Maitri. Journal of Earth System Sciences 116,179–186.

Park, C.G., Dejnakarintra, M., 1973. Penetration of thundercloud electric fields intothe ionosphere and magnetosphere 1. Middle and subauroral latitudes. Journalof Geophysical Research 78, 6623–6633, http://dx.doi.org/10.1029/JA078i-028p06623.

Pawar, S.D., Kamra, A.K., 2000. Comparative measurements of atmospheric electricspace charge density made with the filtration and Faraday cage techniques.Atmospheric Research 54, 105–116, http://dx.doi.org/10.1016/S0169-8095(00)00044-2.

Prospero, J.M., 1984. Aerosol particles in Global Tropospheric Chemistry. NationalAcademy Press, Washington D C.

Raj, P.E., Devara, P.C.S., Maheshkumar, R.S., Pandithurai, G., Dani, K.K., 1997. Lidarmeasurements of aerosol column content in an urban nocturnal boundarylayer. Atmospheric Research 45, 201–216, http://dx.doi.org/10.1016/S0169-8095(97)00037-9.

Takagi, M., 1977. On the regional effect of global electric field. In: Dolezalek, H.,Rieter, R. (Eds.), Electrical Processes in Atmosphere, p. 477.

Fig. 9. Seasonal variation of Potential Gradient (PG) at Kolkata (Lat: 22.561N) alongwith the variation of PG at Potsdam (Lat: 521N) and Johannesburg (Lat: 261S).

S.S. De et al. / Journal of Atmospheric and Solar-Terrestrial Physics 105-106 (2013) 135–141140

Takagi, M., Kanada, M., 1972. Global variation in the atmospheric electric field. Pureand Applied Geophysics 100, 44–50.

Tinsley, B.A., 2000. Influence of Solar wind on the global electric circuit, andinferred effects on cloud microphysics, temperature, and dynamics in thetroposphere. Space Science Reviews 00, 1–28.

Tonev, P.T., Velinov, P.T.Y., 1996. A quasi-DC model of electric fields in theionosphere-ground region due to electrified clouds. Journal of Atmosphericand Terrestrial Physics 58, 1117–1124.

Vojtek, T., Steinbauer, M., Fiala, P., Bartusek, K., 2006. Gerdien condenser measuringand numerical modeling of air ion fields. In: Proceedings of the AppliedElectronics, Pilsen, 6–7 September, 2006.

Williams, E.R., 2009. The global electric circuit: a review. Atmospheric Research 91,140–152, http://dx.doi.org/10.1016/j.atmos.res.2008.05.018.

Williams, E.R., 2013. (personal communication).Wilson, C.T.R., 1920. Investigations on lightning discharges and the electric field of

thunderstorms. Philosophical Transactions A 221, 73–115.Whipple, F.J.W., Scarce, F.J., 1936. Geophys. Mem. vol. 68. Meteorological Office, Air

Ministry, London, UK.Yeboah-Amankwah, D., 1989. Fair weather electric field in Port Moresby. Journal of

Atmospheric and Terrestrial Physics 51, 1035–1040, http://dx.doi.org/10.1016/0021-9169(89)90019-6.

S.S. De et al. / Journal of Atmospheric and Solar-Terrestrial Physics 105-106 (2013) 135–141 141