Analysis and forecasting of precipitation in theLarestan area by Markov chain.

In order to analyze the precipitation of the Larestan area, the rain days with 0.1millimeter or more were obtained from the Iranian Meteorological Organization for the1960-2003 period. First the rainy periods with different lengths were identified andtheir monthly and seasonal frequencies were calculated. On the monthly basis Januaryhad the highest wet days frequency and winter was the wettest but the spring was thedriest season. The wettest year had 44 rain days while only 11 days were experiencedduring the dry year. The mean daily density of rain was 8.2 mm and the mean timeinterval between successive rainy periods was 6.2 days. On the average the rainyperiod begins each year on 8 of December and ends on 6 of April.The first order Markov chain was applied to the data series to forecast the wetperiods. The model responded well and was able to forecast significantly andprecisely. The model was fitted best for the runs of one to six days proving thehypothesis of the study.