Abstract:
Lightning is a fundamental atmospheric phenomenon that significantly affects the Earth’s
climatology. This study uncovers intriguing features and demonstrates strong connections
with other lightning detection networks and instruments. This study examines the detection
efficiency (DE) of World Wide Lightning Location Network (WWLLN) by comparing
WWLLN data to Earth Network Total Lightning Network (ENTLN) data within 50km and
100µs at 20⁰ S to 70⁰ N latitude and 40⁰ E to141⁰ E longitude. Additionally, the variation of
DE due to ENTLN peak current has also investigated in this study. This study also presents
density maps of lightning distribution using 1° by 1° grid boxes and shows a relation with
International Space Station (ISS) Lightning Imaging Sensor (LIS). Diurnal differences in
WWLLN strokes are analyzed across six global continental regions, and lightning activities
over Bangladesh's landmass are assessed. The study employs time series analysis and the
Auto Regressive Integrated Moving Average (ARIMA) modelling technique to develop a
lightning prediction model for WWLLN stroke energy and ISS-LIS flash radiance data,
with standard diagnostic tests evaluating the chosen model's goodness of fit. DE
calculations revealed an average DE of 6.54% (June-July) and 16.61% (NovemberDecember) with 893,773 and 1,393,031 matching WWLLN and ENTLN CGs,
respectively. DE is lowest for peak currents under ±20 kA, increasing from 40% to 70%
above ±50 kA. The maximum DE is 42% for -80 kA and 71% for 100 kA. The mean
positive and negative peak currents for matched ENTLN-WWLLN CG strokes are 48.7 kA
and -44.2 kA, respectively. This analysis reveals that about 60% of the total lightning of
the globe occurs in the landmass and the rest 40% in the oceans. In summer, lightning
maxima align with 30°N, while in winter, they shift towards lower latitudes. The diurnal
amplitude variation for land strokes peaks around 20:00 LT in North and South America,
16:00 LT in Europe, Africa, and Australia, and 14:00 LT in Asia. This study further
highlights that the lightning stroke density (Strokes per km-2 year-1) over Bangladesh is
significantly high and the months of April, May, and June experiencing the highest
lightning activity in the country. Following the necessary diagnostic tests, ARIMA (3, 1,
1) × (2, 1, 0)12 is selected as the best-fitted model for forecasting WWLLN stroke energy,
while ARIMA (2, 1, 2) × (0, 1, 1)12 is chosen for forecasting ISS-LIS lightning flash
radiance.
