Spatio-temporal variability of lightning climatology and its association with thunderstorm indices over India

Lightning is an electrical discharge — a ‘spark’ or ‘flash’ as charged regions in the atmosphere instantly balance themselves through this discharge. It is a beautiful and deadly naturally occurring phenomenon. In June 2020, more than a hundred people died in the state Bihar of India only in 3 days’ span due to lightning events. In this study, Lightning Imaging Sensor (LIS) from the Tropical Rainfall Measuring Mission (TRMM) satellite with a very high spatial resolution of 0.1° has been utilized to create the lightning climatology of India for 16 years during the period 1998 to 2013. Diurnal, monthly, and seasonal variations in the occurrence of lightning flash rate density have also been analyzed. TRMM satellite Low-Resolution Monthly Time Series (LRMTS) with 2.5° resolution datasets have been used for lightning trend analysis. The diurnal lightning event mainly occurs in the afternoon/evening time duration (1400–1900 h) around 0.001 flashes/km2/h. The highest lightning occurred in the month of May (0.04 flashes/km2/day) and the least in December (0.005 flashes/km2/day). The distribution of lightning flash counts by season over India landmass is mainly in pre-monsoon months (MAM) which ranges from 0.248 to 0.491 flashes/km2/day, and monsoon (JJA) ranges from 0.284 to 0.451 flashes/km2/day and decreases afterward. Spatially, the distribution of lightning flash density is seen over the North-Eastern region along with Bangladesh, Bihar, Jharkhand, Orissa, and Jammu and Kashmir region. The pre-monsoon (MAM) season shows a positive trend of lightning around 0.04 flashes in the North-Eastern region and the post-monsoon (SON) shows a negative trend of − 0.021 in central India. The CAPE and K-Index have positively correlated with the flash rate density seasonally but CAPE is more significantly correlated around 0.83 during pre-monsoon and monsoon season. This study also focused on finding out the district wise lightning hotspots regions of India. Rajouri and Reasi district in Jammu and Kashmir has the highest lightning with 121 and 115 flashes/km2/yr, respectively.


Introduction
Lightning is a ubiquitous naturally occurring phenomenon associated with a combination of atmospheric and surface events, including severe weather, which can pose a significant threat to agriculture, electric power networks, property, and living beings (Yair 2018). As a result, in atmospheric science, disaster planning and management, and the power utility/energy sector, the spatial pattern of lightning occurrence is of great interest. Lightning mainly occurs due to thunderstorms which are small synoptic weather events produced due to strong convection (Dowdy 2020). When the collision between the hail particles with other smaller ice particles occurs because this smaller ice particle gains a positive charge and loses a negative charge, respectively. When the charge is built high enough, rapid electricity discharge occurs to equalize the charged region called lightning. (Dwyer and Uman 2014). The lightning frequency depends on the microphysical and kinematics property of the thunderstorms and the available buoyant energy in the environment. It mostly happens in a cumulonimbus cloud, which requires high surface heating, buoyant warm air rises, great heat contrast between surfaces, frontal lifting, or orographic lifting of air parcels (Tapper 1996). Therefore, it can be used to increase our understanding of the structure and variability of thunderstorms at local, regional levels over land and the ocean.
Lightning is one of the most dangerous atmospheric hazards that have an impact on the economy across the globe (Mallick et al. 2022;Mushtaq et al. 2018). The phenomenon of thunderstorm events causes massive destruction and loss of life and property (Dowdy 2020;Bhardwaj and Singh 2017). It occurs more than 3 billion times or 100 times per second (Okafor 2005) all around the earth, and the most prone region for lightning is a tropical region of the earth (Cecil et al. 2014a, b). The range of lightning causalities due to lightning is 6000-24,000 per year, mostly in developing countries than the developed countries (Holle 2016). On June 26, 2020, the lighting event in Bihar caused more than 120 deaths in 3 days. As per the study (Yadava et al. 2020;Singh and Singh 2015;Illiyas et al. 2014), more than 2000 people die due to lightning per year in India, around 9% of deaths due to natural calamities. The most prone lightning area in India is the whole north-eastern region, Bihar, Orissa, Jammu Kashmir, and Uttarakhand. Furthermore, most lightning deaths happen in Maharashtra, Orissa, and Madhya Pradesh due to population density (Yadava et al. 2020).
The cloud-to-cloud flashes are 5-10 times more than cloud-to-ground flashes. Lightning detection from the satellite relies on the sudden release of electrical energy during lightning, generating a rapid heating and shock wave (thunder). The electromagnetic radiation detected by the LIS sensor 128 × 128 charge-coupled device (Goodman et al. 1988) with the viewing area of the sensor is 668 × 668 km with 4.4 km of nadir (Boccippio et al. 2002;Christian et al. 2003). The Lightning Imaging Sensor (LIS) aboard the Tropical Rainfall Measuring Mission (TRMM) produced lightning observations that allowed researchers to examine the electrical, microphysical, and kinematic characteristics of tropical thunderstorms in addition to the hotspot placement over the area (Document describing the (Algorithm Theoretical Basis Document) ATBD-algorithm of the LIS sensor). It covers the whole earth, even areas where there is no ground network coverage, the use of a satellite sensor is a necessity for the study of lightning (Cecil et al. 2014a, b;Cecil et al. 2015;Dewan et al. 2018b;Taszarek et al. 2019;de Abreu et al. 2020;Unnikrishnan et al. 2021). Many studies take place on lightning and its correlation with other indices over the Indian region (Saha et al. 2017;Murugavel et al. 2014;Tinmaker et al. 2014;Kandalgaonkar et al. 2003Kandalgaonkar et al. , 2005Pawar et al. 2012;Ranalkar and Chaudhari 2009;Lal and Pawar 2011;Penki and Kamra 2013). With a rise in CAPE, lightning activity increases as well; in South Asia (> 30 fl /km 2 ), primarily the Himalayan foothills is one of the most lightning-impacted places studied by Albrecht et al. (2016a, b). With the change in CAPE in the foothills of Himalayan, it would change the lightning flash rate by 22% (Ramesh Kumar and Kamra 2012). The lightning prediction has become a difficult area of research. Tiwari et al. (2016) used GCM products to investigate the prediction skill of large-scale seasonal mean temperature fluctuation in the inter annual timescale over India. The performance of each model in simulating NASA's NEX-GDDP-simulated summer monsoon rainfall over homogeneous monsoon regions of India is evaluated using spatial attributes and statistical scores .
The use of a satellite sensor is a vital instrument for the study of lightning. Unnikrishnan et al. (2021), using LIS satellite lightning data, Kandalgaonkar et al. (2005) calculated that a 1° increase in temperature would increase lightning activity by 20-40%. As a result, several kinds of research have been conducted on South Asia to understand better the geographical and temporal variation of lightning and related variables using ground and satellite-based data. During monsoon, CAPE is an essential factor for lightning activities (Murugavel et al. 2014;Pawar et al. 2012). However, there is no relationship between the CAPE and lightning activities in the north-eastern India region due to topographic forcing (Pawar et al. 2010). A study done by Siingh et al. (2014) shows that the CAPE varies all over India, mainly due to orography and vegetation cover. According to a study done by (Pawar et al. (2015), the conversion of moisture in the valley during the night leads to thunderstorms with significant flash rates.
Lightning is linked to thermodynamic variables such as instability, adequate moisture, and lifting force, in addition to human influences (Saha et al. 2012(Saha et al. , 2014Kumar et al. 2016). Although local severe convective storms can occur everywhere in the north eastern region, there are some geographical places where the likelihood of a severe to the strong local convective storm is substantially higher studied done by (Mahanta and Yamane 2020).
CAPE is an index that is used to evaluate atmospheric instability (Romps et al. 2014;Galanaki et al. 2015). The available potential energy transformed into kinetic energy is shown in CAPE, which is used to estimate thunderstorm occurrences; thus, it shows a good correlation with lightning flashes (Galanaki et al. 2015). K-Index is proved to help indicate the air mass thunderstorm. As the K-Index value increases, the probability of a thunderstorm increases (Tinmaker et al. 2017). The K-Index is based on the temperature lapse rate, lower tropospheric humidity, and the vertical extent of the wet layer. A study done by Umakanth et al. (2020) shows that the K-Index value during the premonsoon season shows the intense thunderstorm occurrence over the region Srikakulam district in Andhra Pradesh, suggesting that it provides a good indication in forecasting thunderstorms.
This study shows the spatial and temporal distribution of lightning intensity by estimating flash rate density over the Indian subcontinent. The correlation between lightning intensity and indices such as CAPE and K-Index was investigated seasonally. The lightning trend analysis has also been calculated spatially for the India region seasonally. This study also examined the hotspots region in India with the highest flash rate density that will help us provide early warning and arrangements for the safety of the people, livestocks, and property. Section 2 discusses the study area geography, datasets description, and methodology in details. Next, Section 3 describes the results and discussion part of the research paper, which contains lightning climatology, the correlation between lightning events and indices, lightning hotspots region, and lightning trend analysis. Finally, the conclusion has been described in Section 4.

Study region
India is located in south Asia between 8°4′ north and 37°6′ north latitude and 68°7′ east and 97°25′ east longitude, north of the equator, with a population of 1.21 billion people (2011 census). Figure 1 shows the study area of lightning flash rate density over India region. The Indian Ocean borders India on the south, with the Arabian Sea to the west, the Bay of Bengal to the east, and the Himalayas to the north. According to the Koppen classification, India's climate is split into six primary climate types and microclimates, making it the world's most climatically diverse country; India is primarily a tropical country. The Himalayas and the Thar desert mostly influence India's climate by attracting the monsoonladen southwest wind. The rainfall India receives is around 117 cm, out of which 80% of rainfall is observed during the monsoon months of June to September (Praveen and Talukdar 2020). As per the study, most lightning strikes in India happen in pre-monsoon and monsoon periods, and deaths due to lightning happen in India (Singh and Singh 2015;Yadava et al. 2020, (National Crime Records Bureau) NCRB report).

TRMM satellite datasets
The lightning datasets used in this study are derived from a satellite sensor that is a Tropical Rainfall Measurement Mission (TRMM) which is a National Aeronautics and Space Administration-Japan Aerospace Exploration Agency (NASA-JAXA) cooperative mission with a low inclination equatorial orbit satellite project that is part of NASA's ESE (Earth Science Enterprise). This is the first mission to understand the tropical and subtropical rainfall from a global perspective, which was the least understood parameter. In 1995, NASA launched Optical Transient Detector (OTD), a low earth orbit satellite at 740 km of altitude using a 70° inclination to detect lightning from space with a storm-scale resolution both day and night of 1300 km × 1300 km for each 3 min satellite overpass (Zhang et al. 2019;Boccippio et al. 2002;Christian et al. 2003). LIS combines a high-speed chargecoupled device (CCD) detection array with a narrow band filter centered at 777 nm (Boccippio et al. 2002;Christian et al. 2003) to detect and locate lightning with a storm-scale resolution of 5-10 km, around over tropical region (35 N-S) coverage. For measurement and observation of lightning flash rates for individual storms, the LIS field of view (FOV) in the 90 s (Cecil et al. 2014a, b).
LIS can detect the total lightning throughout day and night with efficiency varying between 69 and 88%, respectively (both within the cloud and from the cloud to the surface) by measuring the radiant energy released by lightning. The overarching goal of LIS was to provide more information on the characteristics of tropical marine and continental convective clouds, laying the groundwork for establishing a worldwide thunderstorm and lightning climatology. TRMM's height was increased from 350 to 400 km in August 2001, resulting in a larger field of vision. It also gives a more extended sample period, increasing the overall number of flashes but not the rate of flashes. Both OTD and LIS are low earth orbit satellites; LIS has a narrower field of view (FOV) than the Optical Transient Detector (OTD) sensor, but it has higher detection effectiveness (Bond et al. 2002), reaching 88% at night (Cecil et al. 2014a, b). The gridded climatology is built with the merging of TRMM-LIS and OTD (Cecil et al. 2014b).
The very high-resolution datasets consist of five gridded data sets: very high-resolution full climatology (VHRFC), very high-resolution diurnal cycle (VHRDC), very high-resolution monthly climatology (VHRMC), very high-resolution seasonal climatology (VHRSC), and very high-resolution annual climatology (VHRAC) were processed by Albrecht et al. (2016a, b). The datasets (daily, monthly, or seasonal) were already treated with a 49-day and 1° boxcar moving average to eliminate diurnal cycle and smooth regions with low flash rates to make the findings more robust.

Methodology
The LIS TRMM (0.1°) very high-resolution gridded lightning climatology data collection (Albrecht et al. 2016a, b) was used to compute the cloud-to-ground lightning flash density over India, to examine the spatial-temporal distribution of lightning in India. Table 1 shows the lightning climatology datasets parameters. The data available in NetCDF format with a resolution of 0.1° over the Indian region was analyzed with the geographic information system (ArcGIS) to examine the flash rate density. As per the earth observation system (EOS/TRMM) data requirements, the seasonal dataset very high resolution seasonal climatology (VHRSC) comprises four seasons i.e. winter (December-January-February), pre-monsoon (March-April-May), monsoon (June-July-August), and post-monsoon (September-October-November). Furthermore, the LIS-OTD (2.5°) low-resolution monthly time series (LRMTS) gridded dataset from 1996 to 2014 has been used (Cecil et al. 2014a, b;Christian et al. 2003).
The CAPE with 2° resolution and K-Index with 0.25° has been used to study the correlation with lightning events. The different spatial resolution datasets have been averaged and interpolated with the coarse resolution for correlation purposes. The lightning hotspot analysis, done by the superimposition of the lightning flash rate density dataset on the district-state map of India, thus finding out the hotspot location district-wise (de Abreu et al. 2020;Albrecht et al. 2016a, b).

Thunderstorm indices
The gridded values of the monthly averaged K-Index (Umakanth et al. 2020;Tinmaker et al. 2017) are among the most significant stability indices for predicting atmospheric convective activity, i.e., CAPE (Dewan et al. 2018a, b;Murugavel et al. 2014;Saha et al. 2017) has been used for the study. The data was obtained from the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA-5 database; ERA5 is monthly averaged data on single levels from 1979 to the present with gridded monthly mean and a spatial resolution of 0.25° downloaded from January 1, 1998 to December 31, 2014(Hersbach et al. 2018). The monthly averaged CAPE with a spatial resolution of 2.0° is obtained from National Centres for Environmental Prediction/National Center for Atmospheric Research (NCEP/ NCAR) reanalysis product (Kalnay et al. 1996) provided by the ESRL PSD, Boulder, Colorado, USA (website: http:// www. esrl. noaa. gov/ psd. html) for the period 1996-2012. CAPE is utilized at the surface level to depict tropical areas more adequately (Tinmaker and Ghude 2019;Dewan et al. 2018a, b;Saha et al. 2017). The monthly mean of the CAPE is utilized to determine the connection with the flash rate density. (Mao and Li 2020;de Abreu et al. 2020).
The following formula is used to calculate CAPE: (where Tv, parcel and Tv, env are the virtual temperatures of the parcel and environment, respectively). The levels of free convection and neutral buoyancy are indicated by the letters Zf and Zn. The CAPE parameter's threshold levels indicate the likelihood of severe weather activity (Grieser 2012). Table 2 depicts the CAPE parameters indicating thunderstorm chances.
The K-Index threshold value indicates the risk of severe weather activity (Johnson 1982). The K-Index may also be used to predict the likelihood of a thunderstorm. It is calculated by subtracting the temperature from the dew point temperature at various atmospheric pressure levels, as shown below (George 1960).
The formula is: Tv, parcel − Tv, env Tv, env dz (2) (T is the temperature in degree Celsius, and Td is the dewpoint temperature at the given pressure level in hPa.); where component [A] shows the static instability between 850 and 500 hPa, component [B] shows the moisture at 850 hPa, and component [c] shows the dryness of the airmass at 700 hPa. Table 3 shows the K index parameters indicating thunderstorm chances.
The index is most effective in flat regions at low to moderate elevations; it is less effective at higher heights. The meaning of the index value changes depending on the season and place (e.g., tropics or mid-latitude, summer or winter). The typical values differ by region; however, it can be advantageous and informative for predicting thunderstorms.

Results and discussions
This section described the spatial-temporal distribution and seasonal variation of CAPE, K-Index, flash rate density, and their trends for 1996-2013 over the India region. This section discussed the lightning hotspot region in India and the correlation between lightning flash rate density with CAPE and K-Index.

Seasonal climatology
The seasonal analysis over the India region, for the period of 1998-2013 has been done by using the very high-resolution seasonal climatology (0. 1°) dataset with the pre-defined seasons from NASA-GHRC datasets, i.e., (winter-(DJF), pre-monsoon-(MAM), monsoon-(JJA), post-monsoon-(SON)). The highest lighting occurs in India during the premonsoon (MAM) and post-monsoon season (JJA) (Yadava et al. 2020). Lightning hotspots in India have more lightning activity (> 15 fl day-1) during the monsoon months (May-October, peaking in September; and less from November to March (Albrecht et al. 2016a, b). As Fig. 2 depicts the lightning density over the India region seasonal wise, thus during the pre-monsoon MAM season highest lightning occurs in India's north-eastern region, around 0.248-0.491 flashes/km 2 /day, wherein West Bengal, Orissa, Bihar, Kashmir, Uttarakhand, and Kerala states which are around 0.129-0.247 flashes/km 2 /day. Due to the low-pressure zone and excessive heat in the north-eastern region of India, including Bangladesh, convection increases the chances of thunderstorms. The moisture availability in the Bay of Bengal, thunderstorm lightning increases in the north-eastern region in the pre-monsoon season (Saha et al. 2017). According to Wu et al. (2016), Backward trajectory analysis confirms that moisture transport from the Arabian Sea and the Bay of Bengal are two critical transport pathways for the summer monsoon intense convective systems at the western end of the southern Himalayan front. In monsoon season (JJA) Kashmir region gets the maximum lightning range in 0.284-0.451 flashes/km 2 /day. During the monsoon season, convection and accompanying lightning occurrences moved from the north-eastern to the northwestern lower Himalayan foothills. According to a study conducted by Saha et al. (2017), the maximum lightning flashes are detected over the north-western section of the Himalayas and much more over the eastern, north-eastern region Himalayas along the foothills of the Himalayas. Cecil et al. (2005), Zipser et al. (2006) and Houze et al. (2007) recognized the western Himalayas as an area with the most severe thunderstorms on earth. Houze et al. (2007) discovered that the most severe convective storms occur upwind or at the mountain foothills, where the moist south westerly monsoon flow from the Arabian Sea meets the descending dry air from the Afghan or Tibetan Plateaus, implying a similarity to dryline convection. Moreover, higher latitudes got more lightning than lower latitudes by Ranalkar and Chaudhuri (2009), which justifies the study entirely.
Furthermore, some parts of Himachal Pradesh and Punjab also get heavy lightning. In post-monsoon (SON), the maximum lightning occurs in Jammu and Kashmir 0.0852-0.143 flashes/km 2 /day, Punjab and Himachal Pradesh around 0.0438-0.0851 flashes/km 2 /day. During the retreating monsoon season, due to western disturbances across northern India. In winter (DJF), the most prone lightning area is northwestern, including Jammu Kashmir, Punjab, 0.0166-0.312 flashes/km 2 /day Himachal Pradesh Haryana, Uttar Pradesh, and some parts of Rajasthan 0.00613-0.165 flashes/km 2 /day. However, the lightning intensity is moderate.

Monthly climatology
The monthly flash rate climatology is examined using the very high-resolution monthly climatology (VHRMC: 1998(VHRMC: -2013 and Low-resolution monthly time series (LRMTS: 1996-2013) datasets. As per Fig. 3, we analyze that the flashes are less during the winter months till March and gradually increase from April and show it peak in May. It is in maximum range during June, July, and August, and after that, it decreases gradually from September. Lightning is sensitive to the surface temperature (Williams 2005), and the study done by Kumar et al. (2016) on the central part and north-eastern part of India analyzed that the increase in land temperature during daytime is one of the regions of lightning in pre-monsoon months.
The overall lighting is high during the monsoon and post-monsoon months. The graph shows the monthly mean flash rate density measured in flashes/km 2 /day. In March, the northeast region, i.e., Meghalaya, Tripura, Assam, got the maximum lightning 0.139-0.294 flashes/km 2 /day, which increases to 0.431-0.752 flashes/km 2 /day in April, and in May, the flashes are around 0.257-0.552 flashes/km 2 /day in the north-eastern region above Bangladesh area cover Assam and also in Kashmir region lightning flashes maximum than rest of India. In August, the lightning flashes are maximum in Kashmir, around 0.253-0.457 flashes/km 2 /day, and in October, the southern part of India, i.e., Tamilnadu, and some parts of Orissa, Bihar, and Kashmir, have the maximum lightning than rest of India. In November, some parts of western ghat and Kerala got lightning. Lightning events occur during September, October and November due to the north-eastern monsoon. The lightning flash count is less than the other months but significant for the area around 0.0308-0.0562 flashes/km 2 /day. Figure 4 demonstrates the mean, maximum, and minimum flash rate density over the India region yearly and

Months
shows the largest flashes occurring in the month of May and gradually decreasing in subsequent months, by using low-resolution monthly time-averaged datasets from 1996 to 2013.

Diurnal climatology
During the daytime efficiency of the LIS, the sensor is less than at night because of the reflection from the top of the clouds and due to sunlight illumination being much brighter than the lightning. The very high-resolution diurnal climatology data (0.1°) is used to calculate the diurnal flashes with the LIS detection efficiency ranging from 69% during the afternoon to 88% at night (Christian et al. 2003;Boccippio et al. 2002;Cecil et al. 2014a, b;Albrecht et al. 2016a, b). The flash rate for the diurnal climatology is calculated in flashes/km 2 /h. Over the India region, the minimum flashes occur during morning time from 0900 to 1100 h, and the peak of the flash detects during the noontime from 1500 to 1900 h of Indian Standard Time (IST) and decreases after that. Figure 5 represents the diurnal flash rate over the Indian region.

Lightning trend analysis
The seasonal trend analysis of flash rate density using lowresolution monthly time series for 1996-2014. LRMTS dataset with the 2.5° resolution is robust for analyzing trends.
The seasonal trend analysis of lightning activity and lightning event spatial distribution has been studied over the Indian region. Due to the wide range of meteorological and environmental conditions and their interactions with topography and landforms, clouds with a wide range of microphysical and dynamical characteristics, thunderclouds forming in different seasons produce a wide range of lightning occurrences in different areas. Lightning activity and its link with pre-monsoon and moist convections, large-scale dynamics, and surface fluxes over the Indian region around the Bay of Bengal and the Arabian Sea have previously been examined by many researchers over the Indian region (Kandalgaonkar et al. 2005;Saha et al. 2014;Chakraborty et al. 2015).
In the winter season DJF, the positive lightning trend is shown in the north-western region of India, including Jammu Kashmir, Rajasthan, Haryana, Uttar Pradesh, Punjab, Uttarakhand, with an increase of 0.001 to 0.004 flashes/km 2 / day and also in the North-eastern region by 0.001 flashes/ km 2 /day. The negative trend is shown in the West Bengal, Bihar, and Orissa region by − 0.008 to − 0.005 flashes/ km 2 / day. In the north-eastern part of India, the MAM pre-monsoon season positive trend is roughly 0.04 to 0.045 flashes/ km 2 /day. J&K has an increase of 0.03 to 0.045 flashes per km 2 /day, whereas West Bengal, Bihar, and Orissa have an increase of 0.025 to 0.035 flashes per km 2 /day. 0.02 to 0.03 flashes/km 2 /day demonstrate a positive trend in Kerala and Karnataka's southern regions. By − 0.005 flashes/km 2 /day, the Kutch region of Gujarat is showing a negative trend. The JJA monsoon season positive lightning trend is shown in J&K around 0.06 to 0.03 flashes/km 2 /day and north-western part around 0.03 to 0.04 flashes/km 2 /day. Chhattisgarh, Orissa, and West Bengal increase by 0.01 to 0.03 flashes/ km 2 /day. There is no negative trend found in the monsoon season in any part of India. There was a positive trend in the SON post-monsoon season in Tamil Nadu, Kerala, Karnataka, around 0.003 to 0.009 flashes/km 2 /day. A negative trend in the north-western part included Rajasthan, Madhya Pradesh, Gujarat by − 0.021 to − 0.015/km 2 /day. Figure 6 clearly shows the spatial trend analysis of lightning over the Indian region.

Correlation
The correlation maps may be quite valuable to obtain regional changes in climate patterns across the Indian subcontinent. Changing climates can deliver mixed benefits and threats to societies, with a disproportional distribution of both. The correlation between lightning flash rate density with convective available potential energy and K-Index will provide an opportunity to investigate the lightning intensity and prime cause related to it further. The spatial correlation analysis was done for 16 years from 1996 to 2012 seasonal basis over the Indian region.

Seasonal spatial correlation between flash rate density and CAPE
As shown in Fig. 7 the spatial correlation between convective available potential energy and lightning flash rate density over the India region seasonally which is positively correlated with each other in the range of − 0.11 to 0.77 during the winter season months DJF. In the range of 0.44 to 0.77, the north-western area, which includes Rajasthan, Haryana, and Punjab, as well as north Jammu & Kashmir, the entire north-eastern region, West Bengal, and Orissa, are positively correlated. Gujarat, as well as several sections of the southern region, like Tamil Nadu and Andhra Pradesh, are negatively correlated. The rest of the area has a moderate degree of correlation. During the pre-monsoon season MAM, the majority region of India, which covers north western, central, and southern India, has a positive correlation in the range of 0.65 to 0.83. In India's north and north-eastern regions, the less correlated region is in the range of 0.076 to 0.048. During the monsoon season JJA, the southern region, and the north-eastern region are significantly correlated in the monsoon season, with correlations ranging from 0.34 to 0.89. Furthermore, Himachal Pradesh, Uttarakhand, Uttar Pradesh, and Jammu & Kashmir have a negative correlation of − 0.66 to 0.16. Only Jammu & Kashmir is adversely correlated around − 0.31 to 0.15 during the post-monsoon SON, otherwise, India as a whole is moderately correlated. The eastern ghat and the north-eastern region have a high correlation of 0.65 to 0.94 during this period.

Seasonal spatial correlation between flash rate density and K-Index
Figure 8 describes the spatial correlation between flash rate density and K index over India and it shows that during the winter season DJF, the region above the tropic of cancer that covers the entire half of India, is positively correlated in the range of 0.43 to 0.83, the highest correlation shown in the eastern ghat, foothills of the Himalayas, and the north-western region of India. The regions which are negatively correlated are Gujarat and Tamil Nadu, around − 0.33 to 0.15. The rest of the region is moderately correlated. In the pre-monsoon season MAM, the majority of India is positively correlated around 0.66 to 0.83. The regions which are negatively correlated are Jammu & Kashmir and North-eastern states around 0.12 to 0.48. During the monsoon season JJA, the Himalaya foothill is negatively correlated to the north-eastern region, around − 0.8 to − 0.021. The north and north-western parts are moderately correlated. Southern India shows a correlation of 0.17 to 0.49. This clearly shows that the correlation between the K-Index and flash rate density during the monsoon season is not good. During the post-monsoon

Hours (IST)
season SON, the region which shows the poor correlation in the western ghat. The rest of India is positively correlated, and the region that shows a high correlation is northern India, the north-eastern region, where the correlation is around 0.79 to 0.93.

Lightning hotspots region of India
The region of lightning hotspots in India presented in this study is above 50 flashes /km 2 /yr. Many studies have been done to present the lightning hotspots over the globe (Albrecht et al. 2016a, b;Cecil et al. 2014a, b), but this analysis will present India's district-wise lightning hotspot regions. The lightning hotspots region is mainly located in Jammu Kashmir, Himachal Pradesh, Uttarakhand, the north-eastern state of India, i.e., Meghalaya, Assam, Tripura west Bengal, Orissa, Jharkhand, Bihar, and Kerala. The foothills of the Himalayas also come under the lightning hotspots. Figure 9 shows the lightning hotspots region of India where most of the lightning hotspots are located in the northern part of India rather than the southern peninsula. Rajouri district in Jammu and Kashmir has the highest lightning activity zone, around 121.40 flashes/km −2 /yr −1 . Below is the list of 50 districts with a lightning hotspot of India where lightning is more than 50 flashes /km −2 /yr −1 , seven districts from Jammu & Kashmir, six from Meghalaya, six from West Bengal, and six from Kerala. There are a total of 52 districts and also the plot to show the lightning hotspot district-wise. Because vast meteorological and environmental factors and their interactions with topography and landforms lead to the creation of clouds with a broad range of microphysical and dynamical features, thunderclouds growing in different seasons show a wide variety of lightning occurrences in different regions. Table 4 mentioned the district-wise list of the top 50 lightning hotspots region of India.

Conclusions
Lightning is one of the most dangerous natural phenomena which causes heavy destruction to human life and property. Lightning mainly occurs in the tropical region of the planet earth. India is situated in the tropical region and due to its location, some of the highest lightning events took place in this region. Because of lightning every year, many human fatalities happen in India. Many types of research have been taking place in the past and still carry on to lessen the impact of lightning. This study has also been done to provide some insights for further research perspectives.
This study has used the satellite datasets from Tropical Measurement Mission Rainfall (TRMM) LIS data and ERA-5 reanalysis datasets, to produce the lightning climatology diurnal, monthly, and seasonally over the India region for 16 years from 1998 to 2013. The spatial plot of seasonal flash rate density shows the region where lightning events are high and low as per the season. The highest lightning events took place in pre-monsoon and monsoon season around 0.49 and 0.45 flashes/km 2 /day, respectively. The affected region in northern India, foothills of Himalaya, and the north eastern region including Orissa and Kerala during these seasons. During the post-monsoon season, lightning strikes are common in the eastern ghats and Kerala around 0.028 to 0.085 flashes/km 2 /day. During the winter, the north western region experienced the most lightning strike ranges in 0.011 to 0.031 flashes/km 2 /day. The north-eastern area, Jammu and Kashmir, West Bengal, Orissa, Bihar, Uttarakhand, Uttar Pradesh, and Kerala are the states with the most lightning strikes.
As per the study in May month, India gets the highest lightning intensity which is around 0.045 flashes/km 2 /day and continues till August and afterward, it decreases gradually. The diurnal variation of lightning intensity is highest during the afternoon from 1500 to 1900 h which is around 0.0012 flashes/km 2 /h and minimum during 0900 to 1200 h in the morning around 0.0002 flashes/km 2 /h. A series of spatial correlation plots were produced seasonally to illustrate essential geographical and temporal properties of lightning with thunderstorm indices, namely convective accessible potential energy (CAPE) and K-Index, to better understand the relationship between indices and flash rate density. The result shows the positive correlation of flash rate density with both the indices but among the two variables, the CAPE is significant with the lightning. During pre-monsoon and monsoon, the correlation of the K-Index is very low which is around 0.49 whereas CAPE is 0.83, respectively. The CAPE is highly correlated in the eastern ghat during the post-monsoon season around 0.94. The CAPE is high in India's eastern, south-eastern, and southwestern regions due to an adequate supply of moisture for the formation of convective instability and thunderstorms. Due to the orographic, the formation of thunderstorms is high in the Northern region of India mainly the foothills of the Himalayas, so lightning events are also more than the rest of India despite less CAPE value over the region (Muhammad et al. 2015). The eastern part of India has a high CAPE value due to the Bay of Bengal sea, which has more moisture for convection than the Arabian sea (Chakraborty et al. 2015). Because India is located in a tropical location, high insolation occurs during the pre-monsoon and monsoon seasons, and owing to the availability of moisture; convective events occur throughout this season. The south-west monsoon wind, which originates in the Arabian Sea and the Bay of Bengal, induces more rainfall from the fairly widespread monsoon trough that runs between western India and the Bay of Bengal (Saha et al. 2014).
The spatial seasonal trend analysis of flash rate density over the Indian region shows a very minimum trend in lightning during the period of climatology. The study provides This study also calculates the lightning hotspots region where the lightning event is more than 50 flashes/km 2 / year in India district-wise. There are more than 50 districts where the lightning is more than 50 flashes/km 2 /year, mostly these regions are located in Jammu and Kashmir, Himachal Pradesh, north-eastern states, West Bengal and Kerala. The Rajouri district in Jammu and Kashmir got the highest lightning around 121 flashes/km 2 /year. This result depicts that the northern region of India is more lightning prone than the southern region. The topography of the region plays the most crucial part in the formation of lightning.