عنوان مقاله [English]
The current study has attempted to estimate and classify the probability occurrence and the return periods of dry and wet years of Northwestern Iran in 55 stations containing 20 years of monthly recorded data (1986-2005), In this regard, at first we evaluated droughts and wet periods of the area by applying SPI index in 8 given ranges as thresholds inside the two-state Markov chain technique. Accordingly, the return periods of each state was calculated and the maps of them were created. The application of Markov chain model indicated that each of the states of intensive wet period, moderate wet period, weak wet period, and weak drought over the study area have followed the second order Markov chain and have maximum probability of occurrence. However, the moderate drought, extreme drought and wetness were not highly supported in Markov chain. The maximum probability of drought occurrence belonged to the southern parts, while the maximum possibility of intensive and weak wet periods have occurred in western and eastern parts. It can be concluded that there is an opposite behavior among these three conditions. In the areas where a possibility of weak and intensive wet periods has stronger probability occurrence, the possibility of intensive drought reduces. In regions where the probability of high moderate wet period occurrence, the intensive drought might mostly happen.