Monthly and seasonal hydrological drought forecasting using multiple extreme learning machine models

Hydrological drought forecasting is a key component in water resources modeling as it relates directly to availability. It crucial managing and operating dams, which are constructed rivers. In this study, multiple extreme learning machines (ELMs) utilized forecast hydrological drought. For purpose, the standardized index (SHDI) precipitation (SPI) computed for 1 3 aggregated months. Two scenarios considered, namely, using SHDI previous months input, SPI input. Considering these two timescales (1 months), 12 input–output combinations generated. Then, five different ELMs support vector machine models used predict on both timescales. preprocessing of data, wavelet hybridized with models, leading 144 models. The results indicate that capable high precision. self-adaptive differential evolution ELM outperforms other has highly positive effect model performance, especially error reduction. general, promising can feasibly be purpose.