The Estimation of Drought and Wet Year Probability by Applying SPI and Marcov Chain (case study: Tehran)

Abstract In this paper, the probability of drought and wet year occurrence in Tehran by applying the first-order Marcov chain on Standardized Precipitation Index (SPI) time series in the scales of 6, 9,12,24 and 48 month were investigated. For this purpose, we used the monthly precipitation data of the Mehrabad station from January 1950 to December 2006. After that, SPI were computed in mention scales. Then, the conformity of SPI time series to the first-order Marcov chain were tested statistically by test statistic X2 and also confirmed. In the first of all, transition probability of drought, wet year and normal state were computed. In addition to, middle number of drought and expected wet year for future were forecasted. The result showed that transition probability of every state to its is more than other states. The probability of Pdd (the probability of drought after drought) is more than Pww (the probability wet state after wet state). Whatever, the time scale of SPI become longer, its stability is increased. On the whole, the probability of normal state is the most among all and the drought equilibrium probability is more than wet year. whatever time scale of SPI became longer, the number of drought is decreased and its duration is increased. In the of all, occurrence probability of drought is more than wet year.