Prediction of Monthly SPI Drought Index over Southwestern Basin of Iran Using Post Processing of CFSv.2 Model

Prediction of drought is not only a main challenge for our country but also is a challenge for those countries with dynamic monthly prediction modeling system as well. In the meteorologically advanced countries, prediction of monthly precipitation through dynamical methods is the first step in prediction of drought indices such as SPI and SPEI. Because southwest watersheds of Iran including Great Karoon, Jarrahi and Zohreh have significant contribution in providing water resources needed in agriculture, drinking, industrial and hydropower, so having a seasonal prediction of precipitation and drought system for management of water resources is of vital importance. In this study, the raw output of CFSv2 seasonal forecast model were calibrated in the base period and then monthly precipitation forecasts for next three months were corrected based on the monthly derived equations. Corrected monthly precipitation were used for three months SPI forecasts. Observed precipitation corrected by APHRODITE dataset and historical forecast were used for calibration. Results showed that, the highest efficiency of drought forecast with 81.5% were occurred in the case of normal precipitation condition of the basin. Efficiency of the drought forecast were 62.3 and 42.3 percent when the precipitation in the basin were in wet or drought conditions. Taking into account the frequency of observed drought events for normal, wet and dry cases which were 76.4, 12.5 and 8 percent, then overall skill of SPI index were found to be 73.4 percent.