Estimating the Standardized Precipitation Evapotranspiration Index Using Data-Driven Techniques: A Regional Study of Bangladesh

Drought prediction is the most effective way to mitigate drought impacts. The current study examined ability of three renowned machine learning models, namely additive regression (AR), random subspace (RSS), and M5P tree, their hybridized versions (AR-RSS, AR-M5P, RSS-M5P, AR-RSS-M5P) in predicting standardized precipitation evapotranspiration index (SPEI) multiple time scales. SPEIs were calculated using monthly rainfall temperature data over 39 years (1980–2018). best subset model sensitivity analysis used determine appropriate input variables from a series combinations involving up eight SPEI lags. models built at Rajshahi station validated four other sites (Mymensingh, Rangpur, Bogra, Khulna) drought-prone northern Bangladesh. findings indicated that proposed can accurately forecast droughts station. predicted better than with lowest mean absolute error (27.89–62.92%), relative (0.39–0.67), (0.208–0.49), root square (0.39–0.67) highest correlation coefficient (0.75–0.98). Moreover, could different scales validation locations. accuracy was for longer periods.