Assessment of Artificial Neural Network through Drought Indices

Prediction of potential evapotranspiration (PET) using an artificial neural network (ANN) with a different architecture is not uncommon. Most researchers select the optimal statistical indicators. However, there still gap to be filled in future applications various drought indices and assessment location, duration, average, maximum minimum. The objective was compare performance PET computed ANN Penman–Monteith technique standardized precipitation index (SPI) (SPEI), two for durations 1, 3, 6, 9, 12–months. Statistical predicted shows RMSE 9.34 mm/month, RSR 0.28, R2 1.00, NSE 0.92, PBIAS −0.04. Predicted based on lower than that approach values higher minimum values. SPEI–Penman–Monteith SPI have monthly correlation greater 0.95 similar severity categories, but SPEI SPI. average prediction show SPEI–ANN captures conditions SPEI–Penman–Monteith. PET–based ANN, performs robustly prediction, fails by degree classification capture when utilized.