Long‐term assessment of ERA5 reanalysis rainfall for lightning events over India observed by Tropical Rainfall Measurement Mission Lightning Imaging Sensor
Prashant Kumar,Shailendra S. Srivastava,Nirav Jivani,Atul K. Varma,Chie Yokoyama,Takuji Kubota
DOI: https://doi.org/10.1002/qj.4719
2024-04-18
Quarterly Journal of the Royal Meteorological Society
Abstract:The European Centre for Medium‐Range Weather Forecasts Reanalysis version 5 (ERA5) reanalysis during lightning events is evaluated over India. The rainfall in the ERA5 reanalysis misses the intensity as well as duration of the peak rainfall at the time of lightning flashes. The ERA5 reanalysis rainfall depicts under‐ (over) estimation of rainfall in plain (orographic) regions. Spatial distribution of (a) terrain height (in meters), and (b) flash density using the Lightning Imaging Sensor over India and surrounding regions (2001–2014). Spatial distribution of total lightning‐associated rainfall (TLAR, in mm·12 hr−1) from (c) GSMaP_ISRO and (d) ERA5 reanalysis during 2001–2014. The ability of numerical weather prediction models to accurately predict extreme weather events, such as thunderstorms marked by heavy rainfall and lightning activities, has consistently been of great importance for human life. The objective of this study is to assess the long‐term reliability of the European Centre for Medium‐Range Weather Forecasts Reanalysis version 5 (ERA5) rainfall in comparison to the Indian gauge‐adjusted Global Satellite Mapping of Precipitation (GSMaP_ISRO) rainfall at the time of lightning flashes measured by the Lightning Imaging Sensor (LIS) onboard the Tropical Rainfall Measuring Mission satellite over the Indian region during the years 2001–2014. This analysis will provide valuable insights into the intricate relationship between lightning flashes and precipitation under various terrain conditions (low, mid, or high), across oceanic regions (Bay of Bengal, Arabian Sea), and during different monsoon phases (normal, active, or deficit). According to a prolonged examination of LIS data, April–June accounts for ~50% of the total flashes, with the largest number of flashes occurring over the Himalayan and the northeastern part of India. According to hourly GSMaP_ISRO rainfall, the most substantial lightning‐associated rainfall happens an hour prior to lightning flash (T − 1) and within three hours after (T + 3), indicating a robust correlation between heavy rainfall and lightning activity during this time frame. The rainfall in the ERA5 reanalysis misses the intensity as well as duration of the peak rainfall at the time of lightning flashes. Furthermore, the ERA5 reanalysis rainfall depicts under (over)‐estimation of rainfall in plain (orographic) regions. The underestimation of ERA5 rainfall is very pronounced over the Indian Ocean and Bay of Bengal regions, mainly between flash time (T) to two hours after the flash time (T + 2). The results indicate that there is a requirement for additional enhancements in the ERA5 reanalysis rainfall for lightning occurrences.
meteorology & atmospheric sciences