Journal of Atmospheric and Solar-Terrestrial Physics 120 (2014) 143–149

Contents lists available at ScienceDirect

Journal of Atmospheric and Solar-Terrestrial Physics

http://d1364-68

n CorrE-m

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

Atmospheric surface layer responses to the extreme lightning day inplateau region in India

Arun K. Dwivedi a,b, Sagarika Chandra b,n, Manoj Kumar c, Sanjay Kumar d,N.V.P. Kiran Kumar e

a Department of Electrical & Electronics Engineering, Birla Institute of Technology, Mesra, Ranchi 835215, Indiab Centre of Excellence in Climatology, Birla Institute of Technology, Mesra, Ranchi 835215, Indiac Centre for Environmental Sciences, Central University of Jharkhand, Ranchi 835205, Indiad Department of Electronics & Communication Engineering, Birla Institute of Technology, Mesra, Ranchi 835215, Indiae Space Physics Laboratory, VSSC, ISRO, Thiruvananthapuram 695022, India

a r t i c l e i n f o

Article history:Received 29 May 2013Received in revised form9 August 2014Accepted 9 August 2014Available online 19 August 2014

Keywords:Atmospheric surface layerIrradianceSensible heat fluxStability parameter

x.doi.org/10.1016/j.jastp.2014.08.00326/& 2014 Elsevier Ltd. All rights reserved.

esponding author.ail address: chandrasagarika7@gmail.com (S. C

a b s t r a c t

This paper discusses the observations of the atmospheric surface layer (ASL) parameters during thelightning event. During this event behaviour of surface layer parameters has been observed. Otherderived parameters like Monin–Obukhov stability parameter (z/L), turbulent kinetic energy (TKE),momentum flux (MF) and sensible heat flux (SHF) have also been considered during this stochasticphenomenon. Characteristics of these surface layer parameters have been analysed during lightningperiod and compared with the clear weather day. During the peak period of the lightning, the incomingsolar irradiance was reduced by one third of its normal value, resulting in an air-temperature decrementnear the surface in the range of 4 °C to 6 °C. In addition to that a significant reduction in energyexchanges between surface and lower lying atmosphere (viz. TKE, MF and SHF), has also been observed.The rate of instantaneous decay in solar irradiance and SHF from the first strike to its peak strike timewas larger than that seen during clear day hours. The normalized standard deviations of windcomponents during clear day were studied using Monin–Obukhov similarity theory (MOST) and theresults have been compared with earlier studies reported in the literature.

& 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The atmospheric surface layer is an active link between theatmosphere and the surface of the earth. Thus, its ability totransport momentum, sensible heat, water vapour and otherconstituents is of fundamental importance in all studies relatedto surface–atmosphere exchange process (Stull, 1988). The char-acteristics of this layer are highly influenced by convection,turbulence, thunderstorm, lightning and its interaction with thefree atmosphere. In addition, incoming solar radiation plays animportant role in various surface layer processes. As the solarradiation is the primary source of energy for living being on theearth and plays an important role for the exchange processesbetween earth surface and overlying atmosphere due to radiativeheating and hence affects the variation in atmospheric tempera-tures. As the day progressed surface gets heated due to theintensification of solar radiation incident over the surface. The

handra).

maximum heating is observed during mid-noon when the turbu-lent exchange processes are well mixed by that period. Lightninghas been considered as a phenomenon in which solar irradiancebecomes very less due to cloud cover and it develops to the usualvalue. Later it becomes very less and eventually follows the usualcurve. This situation has been taken as a required condition tostudy the response of ASL in the absence and the availability ofsolar radiation in a very small time span.

The event of lightning associated with thunderstorm has al-ways been very useful to meteorologists for studying the responseof the atmosphere. Lightning is a flash of light in the atmospherewhich consists of thunder and is caused by a discharge ofelectricity. It has been estimated that lightning strikes on theearth surface about 100 times every second (Srinivasan and Gu,2006). The sudden heating effect and the expansion of heated airgive rise to a supersonic shock wave in the surrounding air, whichis heard as thunder (Rakov and Uman, 2003). Lightning isgenerated in electrically charged storm systems; however themethod of cloud charging still remains elusive. When the electricfield becomes strong enough, an electrical discharge occurs withinthe clouds or between clouds and the ground. In this analysis, we

Response

time

Accuracy

o1min

710

%o

5s7

0.1°C

12–15

s7

1.5%

2s

70.5hPa

o5s

70.3m/s

o5s

73°

Samplingrate

0.1–

10Hz

0.1–

10Hz

Accuracy

71%

A.K. Dwivedi et al. / Journal of Atmospheric and Solar-Terrestrial Physics 120 (2014) 143–149144

have investigated the impact of this high electric field on thecharacteristics of the ASL over the Chota Nagpur Plateau area(Ranchi). Results obtained might facilitate in understanding thedecay and growth of turbulence process during lightning periodadditionally showing the response of ASL parameters. The plateauregion has a strategic location from the point of view of pre-monsoon thunderstorm evolution. Lightning is a common phe-nomenon during pre-monsoon thunderstorm and also duringactive monsoon phases. Every year a lot of casualties have beenobserved over the region due to lightning and therefore stategovernment agencies have also started working to understand thisphenomenon over the region. Looking at the importance of thisphenomenon, it has been tried to understand the surface layerprocesses during pre-monsoon lightning period in this paper. Thispaper also presents the influence of lightning on turbulent andfluxes parameters. Therefore, objectives of the paper are to study(1) the variations in the ASL parameters during lightning and clearday, (2) the turbulent characteristics of wind components for clearday according to Monin–Obukhov similarity relations.

Table

1Details

ofinstrumen

tsusedon

thelig

htningday

aswellas

clea

rday.

"\wideh

attbch

mod

eval"M

ax(a)Automatic

wea

ther

station(M

ake:

Cam

pbe

llScientific,

Can

ada)

Nam

eofth

esenso

r(m

easu

redparam

eter)

Ran

ge

CM3py

ranom

eter

(solar

radiation

)30

5–28

00nm

(50%

points)(spectral

range

)Pt10

0RTD

(air

temperature)

�40

–60

°CHyg

roclip-S3(Relativehumidity)

0–10

0%Electron

icba

rometer

(atm

ospheric

pressure)

600–

1100hPa

Anem

ometer

(windsp

eed)

0–10

0m/s

Windva

ne(w

inddirection

)0–

360°

(b)So

nic

anem

ometer

(Mak

e:Cam

pbe

llScientificparam

eter,C

anad

a)

Nam

eofth

esenso

r(m

easu

redparam

eter)

Ran

ge

Ultra-son

ican

emom

eter

(u,v

andw)

0–10

0m/s

(0–22

4mph)

Air

temperature

(T)

�40

to60

°C(c)Electric

fieldmeter

(Mak

e:Cam

pbe

llScientific,

Can

ada)

Nam

eofth

esenso

r(m

easu

red

param

eter)

Ran

ge

Electric

fieldmeter

AEF

7(0–21

,000)

2. Data and methods of analysis

In order to analyse the impact of lightning on general char-acteristics of the atmospheric surface layer over Ranchi, the timeseries measurement of different meteorological parameters (i. e.incoming solar radiation, air temperature, relative humidity, pres-sure, wind speed, etc.), were made. Wind direction was operatedin affiliation with 3-D fast response sonic anemometer (fixed at10 m above ground level) for measuring wind speed in threedifferent axes and temperature at high frequency. Surface layermicrometeorological observations were obtained from the auto-matic weather station (AWS) at 8.5 m level which were averagedfor one hour interval. The fast response measurements of windspeed and temperature were recorded from the sonic anemometerat a frequency of 10 Hz for the determination of TKE, MF, and SHF.For the collection of Atmospheric Electric Field (AEF) data, anelectric field meter (EFM) has been installed. It measures at onesecond interval which is averaged for one minute duration. TheAEF has been measured for decades by EFM usually known as“field mills”. Traditional field mills employ a spinning metal rotor(vane) electrically connected to Earth ground, placed between theexternal field and stationary metal sense electrodes. Details ofinstruments used on the lightning day as well clear day are givenin Table 1.

The surface layer characteristics are defined in terms ofstability, and the vertical variation of temperature, relative hu-midity, wind speed, fluxes and turbulence. The magnitude ofturbulence within the surface layer is estimated through themagnitudes of TKE, while the turbulent exchange of momentumand heat from the basic surface are measured through the verticalmomentum flux (τ) and sensible heat flux (HS). These parametersare determined by (Stull, 1988):

= ′ + ′ + ′( )u v wTKE12 (1)

2 2 2

τ ρ= ′ ′ + ′ ′( )u w v w (2)2 2

ρ θ= ′ ′H C w( ) (3)S P

In the above equations ρ and CP are the air density and specificheat of air. u', v', w' and θ′ are the fluctuations from the mean ofzonal, meridional, vertical winds and the potential temperaturerespectively. The bar over the equation indicates the value aver-aged in 30 min.

A.K. Dwivedi et al. / Journal of Atmospheric and Solar-Terrestrial Physics 120 (2014) 143–149 145

Added to these parameters, the stability of the surface layerwas characterized through z/L. The negative and positive values ofz/L represent unstable and stable conditions while z/L¼0, manifestneutral conditions in the surface layer. Here, z (¼10 m) is themeasurement height and L is the Obukhov length, defined withthe help of friction velocity as follows (Stull, 1988):

κθ

θ= − ′ ′⁎L ug

w/(4)v

v3

⎛⎝⎜

⎞⎠⎟

where

τ ρ=⁎u ( / ) (5)

where κ is the Von-Karman constant (¼0.4) and g and θvrepresents the acceleration due to gravity and virtual potentialtemperature respectively.

To compare our turbulent measurements for the clear day withmicrometeorological measurements reported earlier, Monin–Obu-khove similarity theory (MOST) has been applied. According toMOST, in the ASL various parameters become universal function ofstability parameter (z/L) when normalized by appropriate scalingparameters. In this theory, dimensional analysis has been used toobtain empirical relation between certain dimensionless variables.There are various experiments have been done to verify MOST. Ithas been shown that MOST has not always been successful. Mahrt(1998) has shown that similarity theory breaks down into verystable regime. It also breaks down for horizontal velocity compo-nents during unstable and convective conditions.

Fig. 1. Atmospheric Electric Field during lightning day. (a) Diurnal variation.(b) High resolution during 1000–1500 hrs.

3. Results

3.1. Surface layer characteristics over the plateau region duringlightning period

The surface layer of the atmospheric boundary layer respondeddifferently. The remarkable differences in its characteristics duringlightning day and clear day were observed. Here, surface layermeteorological observations were made on 30 May 2011 (lightningday) and compared with the mean of similar observations re-corded on 06 April 2011 (clear day) in the same pre-monsoonseason. Measurements corresponding to cloudy days are excludedfrom the analysis. In clear day, the maximum irradiance rangedbetween 350W m�2 and 832 W m�2 with a mean value of 640W m−2. The lightning events are marked as part A and B as shownin Fig. 1(a). The high resolution plot for lightning period during1000 hrs to 1500 hrs has been shown in Fig. 1(b). The negativepolarity of electric field is expected for cloud to ground dischargesat any range, since they effectively remove the negative chargefrom the ground (Rakov and Uman, 2003).

The temporal variation of the irradiance has been shown inFig. 2(a). It is clear from the figure that it has a maximum value at1100 hrs local time on the clear day; however the diurnal patternof irradiance on 30 May 2011 (lightning day) is very muchdifferent during the period of lightning. Before the first stroke ofthe lightning at 1000 hrs local time, the intensity of irradiance isreduced and the magnitude of irradiance decreased to 295 Wm�2.This rate of reduction in irradiance is almost one third time thanthat of during clear day which shows the rapidity of the event. Theirradiance shows a rapid recovery and attained a magnitude of812 W m�2 at 1100 hrs local time. Again it goes down just beforethe second strike. After that it reaches to its normal value whichresembled the clear day.

Energy fluxes show some remarkable changes during thelightning. In response to the reduced irradiance during this period,the magnitudes of sensible heat flux and air temperature also

shows a similar signature as seen in the irradiance (Fig. 2(e) and(b)). On the lightning day, the sensible heat flux attained its peakmagnitude (E144 W m�2) at the 1200 hrs of local time; howeverdue to the presence of cloud the incoming radiation is low so thatthe magnitude of sensible heat fluxes is also very low �58 W m�2

as shown in Fig. 2(e). Then it increases to the maximum value justafter the lightning and again suddenly comes down just before thesecond strike. However, in case of the clear day no abrupt changesoccurred during that period. Air temperature ranged from 16 °C to32.2 °C on the clear day with its normal diurnal peak in theafternoon time 1400 local time. In case of lightning day themagnitude of temperature is found to be different and observedby 2 to 4 °C higher than that of the clear day as shown in Fig. 2(b).The rate of decrease in air temperature in the late evening hoursafter 1800 hrs on the lightning day is not relatively sharper thanthe clear day. Temporal variations in air temperature are oftenrelated to the variation in relative humidity as they are inverselyproportional to each other. In the present investigation, relativehumidity variations also showed a typical diurnal pattern withhigh values which were opposite to the air temperature variations.Interestingly, relative humidity values after the lightning periodincreases very smoothly as shown in Fig. 2(d).

TKE shows a significant rise at about 1000 local time, indicatingstrengthen the turbulence in response to the lightning event andremain coarsely changes during the whole period as shown inFig. 2(f). A sharp rise in TKE (�5 m2 s�2) has been observed duringthe peak period of lightning. On the clear day, TKE is found to be1.5 m2 s�2 at 1000 hrs whereas on the lightning day it increasedup to 3 times approximately. This is attributed to the highturbulent structure of the atmosphere due to atmospheric in-stability prevailed.

Surface layer wind speed for the clear day was normal andranged between 0 to 2.5 m s�1 with the diurnal peak in the mid-afternoon at about 1400 local time. Wind speed variations on

Fig. 2. Diurnal variability in atmospheric surface layer parameters on the lightning day compared with the normal clear day: (a) incoming solar radiation; (b) air temperature(AT); (c) momentum flux (MF); (d) relative humidity (RH); (e) sensible heat flux (SHF); (f) turbulent kinetic energy (TKE); (g) wind speed and (h) Monin–Obukhov stabilityparameter (red colour corresponds to clear day and black colour represents lightning day). (For interpretation of the references to colour in this figure legend, the reader isreferred to the web version of this article.)

A.K. Dwivedi et al. / Journal of Atmospheric and Solar-Terrestrial Physics 120 (2014) 143–149146

A.K. Dwivedi et al. / Journal of Atmospheric and Solar-Terrestrial Physics 120 (2014) 143–149 147

lightning day are quite random in nature which shows the abruptchanges (1.4 m s�1 to 1.8 m s�1) as shown in Fig. 2(g). Thevariations of momentum flux were almost in tune with windspeed and showed a sudden decrease during lighting period. Themomentum flux has been observed to rise for a few hours beforethe lightning event due to arrival of strong wind. However, theformation of cloud condensation with the available moisture justbefore the lightning event causes drastic reduction of momentumflux. A sharp rise in momentum flux is observed which increasedto a maximum value at � 1000 hrs of the local time and suddenlydecreases during lightning time as shown in Fig. 2(c). Then it goesto the normal value and again comes down in the second strike.During the peak noon time, the magnitude of the momentum fluxon the clear day was about 5.9 N m�2 while it is 2 N m�2 on thelighting day. Though, the wind speed and momentum flux variedconsiderably on the clear day during the period 0800–1500 localtime, a sudden decrease in their magnitude at about 1100 and1400 local time in the lightning day as shown in Fig. 2(c) and (g).The surface layer momentum flux has shown a peak duringlightning period. A sudden increase of wind speed during thisperiod resulted in the presence of high momentum flux within thesurface layer. The mean wind speed affects surface layer para-meters such as MF and TKE. The variation of MF and TKE during

Fig. 3. Normalized flux parameters. (a) Momentum flux (MF), (b) Turbulent kineticenergy (TKE) (red colour corresponds to clear day and black colour representslightning day). (For interpretation of the references to colour in this figure legend,the reader is referred to the web version of this article.)

lightning and clear day after normalizing them by the square ofmean wind speed, is shown in Fig. 3(a) and (b) respectively.

On the clear day, z/L is found to vary between �0.5 and 2,indicating moderately unstable to stable conditions while thisfeature is totally different in the lightning day. The value of z/L hasbeen �7.9 just a few hours before the first lightning strike whichshows the highly unstable condition. It goes near to the neutralcondition again just after this event and falls in the negative valuebefore the second striking period of lightning. After the secondlightning event, it goes to the value which is quite equivalent tothe clear day. Total rainfall for the lightning day is shown in Fig. 4.It is clear from the figure that there is rainfall just after the firstlightning strike which may be responsible for the lowering themagnitude of the turbulent and flux parameters. The lower valueof these parameters can be attributed to the fact that the strengthof second lightning strike is less as compared to the first strike.

3.2. The non-dimensionalized wind components as a function ofstability parameter

The normalized standard deviations of wind components u, vand w plotted against z/L have been show in Figs. 5–7 respectively.As shown in Fig. 5, most of the data points lie in stable or nearneutral region. The value of su/un increases with increase in z/L forstable condition. This behaviour has also been verified by otherresearchers (Hsieh and Katul, 1997; Pahlow et al., 2001). Theturbulent motion is damped for this very stable regime becauseof very less value of un. Therefore the stable condition is prominentfor increasing z/L for u component. As like for su/un, the normal-ized standard deviation for v component sv/un follow the similarpattern. The value of sv/un is also increasing with increase in z/Lbut not much as su/un. There is a lot of literature published forsw/un (Chu et al., 1996; Mahrt, 1998; Krishnan and Kunhikrishnan,2002; Moraes et al., 2005) while only few is available for su/un andsv/un. However the behaviour of sw/un is well known but it can bechanged according to different types of terrain. It shows someuncertainties when plotted against z/L. The empirical relation forthe u, v, and w components has been given below:

σ⁎

= + ⁎iu

a b z L( / )(6)

c

where i represent wind components u, v, w respectively and a, b, care constants whose values have been given in Table 2.

Fig. 4. Total rainfall during the lightning day.

Fig. 5. Normalized standard deviation of longitudinal wind velocity as a function of z/L for (a) unstable and (b) stable conditions.

Fig. 6. Normalized standard deviation of lateral wind velocity as a function of z/L for (a) unstable and (b) stable conditions.

Fig. 7. Normalized standard deviation of vertical velocity as a function of z/L for (a) unstable and (b) stable conditions. The solid lines represent the empirical relationsapplied to the present data.

Table 2Coefficients of the fitting curves for the normalized standard deviation of the windcomponents.

u v w

a b c a b c a b c

Stable 0.7 3.2 0.24 0.7 2.1 0.1 0.2 1 0.34Unstable 2.1 �3.45 0.65 0.9 �18.5 0.6 0.75 �3 0.65

A.K. Dwivedi et al. / Journal of Atmospheric and Solar-Terrestrial Physics 120 (2014) 143–149148

4. Discussion and conclusion

The event of lightning associated with thunderstorm has al-ways been useful to the meteorologists for evaluating the responseof the atmosphere. A number of studies have been conductedduring thunderstorms that included observations of meteorologi-cal parameters such as air and soil temperature, solar irradiance,humidity, wind speed and turbulence. Many of these studies havereported significant changes in the mean meteorological para-meters within the surface layer, associated with the thunderstormevent (Chaudhuri and Middey, 2013; Grant, 1986). The turbulentstructure analysis in the convective boundary layer has beeninvestigated (Kaimal et al., 1976). Kang (2009) conducted anexperiment to examine the temporal oscillations in the convectiveboundary layer by mesoscale surface heat flux variations. Beringerand Tapper (2002), during the maritime continent thunderstormexperiment (MCTEX) conducted on the Tiwi Island, NorthernAustralia, measured the surface energy exchange above the domi-nant surface types. Grant (1986), made an observation under the1981 KONTUR experiment obtained turbulence data throughout

the depth of the mixed-layer in near neutral to slightly convectiveconditions. Spectral analysis is used to investigate the variety ofturbulence length scales with height and mixed layer eddystructure. Manohar et al. (1999), made some observations ofthunderstorm in different seasons of the year, they describe themonthly mean electrically conditions of isolated deep convectivestorms at Pune, India and noted that the electrification of the premonsoon season thunderstorms dominated by a factor of 3–4 overthe monsoon seasons.

A.K. Dwivedi et al. / Journal of Atmospheric and Solar-Terrestrial Physics 120 (2014) 143–149 149

In comparison to the earlier reported observations, resultsobtained from the present study additionally offers supportingproof in favour of instant cooling of the air, a marginal increase inrelative humidity and the decay in turbulence intensity through-out the lightning. Real atmosphere is neither utterly dry norsaturated. It remains somewhere in between them. Just beforethe prevalence of lightning the atmosphere becomes terribly dry.Surface layer responds differently on lightning and clear weatherconditions. The wind speed plays a very acute role at this station,Ranchi. The study shows that the surface layer becomes turbulentones the mechanical turbulence dominates because of high windspeed throughout the lightning and thunderstorm event. Thistransition between turbulent and stable physical phenomena isresponsible for the changes in several stability parameters. Achange in atmospheric stability from unstable to stable conditionsduring three phases of the lightning and in which z/L variedbetween �7.9 to 0.01. To verify the turbulence measurements forclear day with earlier studies, the normalized standard deviationof all wind components were analysed according to MOST. It wasfound that su/un, sv/un, and sw/un obey similarity theory understable and unstable conditions. The su/un and sv/un componentswere well scattered and shown variability when plotted againstz/L. The sw/un component obeyed the 1/3 power law underunstable condition which has also reported earlier by Panofskyet al. (1977) for flat terrain and Founda et al. (1997) for complexterrain. The mean normalized standard deviation of vertical windvelocity is found to be 1.26 which is quite near to 1.37 and 1, thevalues found by Krishnan and Kunhikrishnan (2001) for Ahmeda-bad region and Ramana et al. (2004) for Lucknow region respec-tively. It is suggested that the increase in si/un is due to themesoscale phenomena which are non-turbulent (Högström, 1990).The results by Mahrt (1998) also support the hypothesis. Apartfromwind components TKE also shows a significant change duringthis type of convective event. TKE increases directly with theadvancement of the electrical storm towards the station. Obser-ving the different TKE and surface energy fluxes, or the temporalchanges of various surface layer's parameters are very useful infurther studies like boundary layer modelling and lightning andthunderstorm forecasting over this area.

Acknowledgement

Authors are thankful to Department of EEE, BIT Mesra forproviding Electric field data under DST, GoJ, Project. Authorsthankfully acknowledge MoES – GOI for providing support underits mega field campaign (Grant no. MoES/568/2011-2012/CASH).Authors are also thankful to the Director, Space Physics Laboratory,Thiruvananthpuram for providing necessary support under NOBLEproject (Grant no. SPL/BLPAM/BITS/10) and MBLM data was

utilized from ISRO’s PRWONAM Project (Grant no. E.33015/5/2011).

References

Beringer, J., Tapper, N., 2002. Surface energy exchanges and interactions withthunderstorms during the Maritime Continent Thunderstorm Experiment(MCTEX). J. Geophys. Res. 107, http://dx.doi.org/10.1029/2001JD001431 (AAC3-1–AAC 3-13).

Chaudhuri, S., Middey, A., 2013. Study of near surface boundary layer characteristicsduring pre-monsoon seasons using micrometeorological tower observations.Atmósfera 26, 125–144.

Chu, C. R, Parlange, M.B., Katul, G.G., Albertson, J.D., 1996. Probability densityfunctions of turbulent velocity and temperature in the atmospheric surfacelayer. Water Resour. Res. 32, 1681–1688.

Founda, D., Tombrou, M., Lalas, D.P., Asimakopoulos, D.N., 1997. , Some measure-ments of turbulence characteristics over complex terrain. Bound.-Layer Me-teorol. 83, 221–245.

Grant, A.L. M., 1986. Observations of boundary layer structure made during the1982 KONTUR experiment. Q. J. R. Meteor. Soc. 112, 825–841. http://dx.doi.org/10.1002/qj.49711247314.

Hsieh, C.I., Katul, G.G., 1997. Dissipation methods, ’Taylor's hypothesis, and stabilitycorrection functions in the atmospheric surface layer. J. Geophys. Res. 102,16391–16405.

Högström, U., 1990. Analysis of turbulence structures in the surface layer with amodified similarity formulation for near neutral conditions. J. Atmos. Sci. 47,1949–1972.

Kaimal, J.C., et al., 1976. Turbulence structure in the convective boundary layer. J.Atmos. Sci. 33, 2152–2169 (doi:10.1175/1520-0469(1976)033⟨2152:TSITCB⟩2.0.CO;2).

Kang, S.L., 2009. Temporal oscillations in the convective boundary layer forced bymesoscale surface heat-flux variations. Bound.-Layer Meteorol. 132, 59–81.http://dx.doi.org/10.1007/s10546-009-9391-5.

Krishnan, P., Kunhikrishnan, P.K., 2002. Some characteristics of atmospheric surfacelayer over a tropical inland region during southwest monsoon period. Atmos.Res. 62 (1), 111–124.

Mahrt, L., 1998. Stratified atmospheric boundary layers and breakdown of models.Theor. Comput. Fluid Dyn. 11, 263–279.

Manohar, G.K., et al., 1999. Thunderstorm activity over India and Indian south-westmonsoon. J. Geophys. Res. 104, 4169–4188. http://dx.doi.org/10.1029/98JD02592.

Moraes, O.L.L., Acevedo, O.C., Degrazia, G.A., Anfossi, D., da Silva, R., Anabor, V.,2005. Surface layer turbulence parameters over a complex terrain. Atmos.Environ. 39 (17), 3103–3311.

Pahlow, M., Parlange, M.B., Porté-Agel, Fernando, 2001. On Monin–Obukhovsimilarity in the stable atmospheric boundary layer. Bound.-Layer Meteorol.99 (2), 225–248.

Panofsky, H.A., Tennekes, D.H., Wyngaard, J.C., 1977. , The Characteristics ofturbulent velocity components in the surface layer under convective condi-tions. Bound.-Layer Meteorol. 11, 355–361.

Rakov, V.A., Uman, Martin A., 2003. Lightning Physics and Effects. CambridgeUniversity Press, Cambridge, England.

Ramana, M.V., Krishnan, P., Kunhikrishnan, P.K., 2004. Surface boundary-layercharacteristics over a tropical inland station: seasonal features. Bound.-LayerMeteorol. 111 (1), 153–157.

Srinivasan, K., Gu, J., 2006. Lightning as atmospheric electricity. In Proceedings ofthe Canadian Conference on Electrical and Computer Engineering, pp: 2258-2261, doi: 10.1109/CCECE.2006.277659.

Stull, R.B., 1988. An Introduction to Boundary Layer Meteorology. Kluwer AcademicPublishers, Dordrecht.