Physical and Dynamical Characteristics of Thunderstorms Over Bangladesh Based on Radar, Satellite, Upper-Air Observations, and WRF Model Simulations

Abstract

The physical and dynamical characteristics of pre-monsoon (March–May) thunderstorms (TS) are investigated over Bangladesh using the Weather Research and Forecasting (WRF) model. The model was run for 24-h using 6-hourly data sets as initial and lateral boundary conditions. The Milbrandt cloud microphysics scheme, Krain–Fritsch cumulus scheme, and Yonsei University planetary boundary layer are found to be the best combination by comparing the root mean square error of rainfall from 18 sensitivity experiments. The synoptic condition and atmospheric instability associated with three TS cases have been analyzed based on mean sea level pressure, convective available potential energy (CAPE), and thermodynamic indices. CAPE is found between 500 and 2700 J kg−1 at 0000 UTC for the TS cases. The presence of low-level southerly wind from the Bay of Bengal and upper-level westerly or north-westerly wind with speeds ranging from 5 to 30 m s−1 is found during the storms. The cloud characteristics and intensity were investigated based on Doppler Weather Radar and INSAT 3DR satellite observations. Cloud micro-physical hydrometeors and other composite features (e.g., cloud top and core precipitation altitude, length of squall line) were also studied by the model and then compared with available observed values. For 24-h rainfall, 2\u2009×\u20092 contingency tables are computed using Model Evaluation Tools. Categorical skill scores such as proportion correct, frequency bias index, probability of false detection, false alarm ratio, and Gilbert skill score were calculated to evaluate the model s performance. The forecast goodness is found to be reasonably satisfactory after analyzing the significance of the skill scores.