Exploring the Thunderstorm Predictors in Indonesia

Abstract

Indonesia, as a tropical region and prone to frequent thunderstorms, is faced with potential risks to both human life and infrastructure. The analysis of a dataset containing observations of thunderstorms enabled us to identify variations in the vertical air temperature profiles between days experiencing thunderstorms and those characterized by no significant weather conditions (Nosig) within the layer between 1000 and 150 hPa. Examination of the relative humidity in the middle-troposphere during the thunderstorm days exhibited elevated moist layers in compared to the Nosig days across all the investigated regions. By employing stability indices that reflect the atmospheric conditions favorable for thunderstorm development, in conjunction with gridded reanalysis data and logistic regression methodologies, we ascertained that among the numerous convective stability parameters scrutinized, precipitable water (PW), K index (KI), and relative humidity in the middle troposphere (RH_Middle) emerged as the most effective parameters in distinguishing between environmental conditions conducive to thunderstorms and those devoid of significant weather phenomena. Assessment utilizing receiver operating characteristic curves illustrated that the optimal normalized thresholds for PW, KI, and RH_Middle were 0.67, 0.86, and 0.62, respectively.