Forecasting the Reconnaissance Drought Index (RDI) Using Artificial Neural Networks (ANNs)

The study of drought and its forecasting plays an important role in planning managing water resource systems, especially extreme climatic periods. This aims to analyze forecast characteristics, through the use Reconnaissance Drought Index (RDI) order temporal spatiotemporal nine climate stations Kurdistan Region Iraq for period (1973-2020) detect beginning end period, as well future droughts using two types artificial neural networks: Recursive Multi-Step Neural Networks (RMSNN) Direct Network (DMSNN). results revealed that driest years were (1998-99) Amadiyah, Erbil Sulaymaniyah stations, (2007-08) rest area. Moreover, models depending on simulation methods adopted have shown ability these with regard last six years, both increase amount error we go forward. However, (DMSNN) model was more accurate, by statistical tests.