Abstract

Abstract―This study evaluated the probable length in days of dry and wet spells in Khuzestan province using daily rainfall data from 11 synoptic stations during the period 1990–2012. Transfer matrix calculations and model tests were performed using Matlab software. For interpolation maps, kriging in ARC GIS 10.2 software was used. Results indicated that the probability of dry days occurring in Khuzestan province increases from the south to the north. The highest probabilities of a two-day dry spell were 15% in the north and 12% in the northeast of the province. The probability of a 3-day dry spell in the north and northeast was 10%. Moving to the south and southwest of the province, the probability decreases, amounting to 6% in the southwest. Results of calculations of return periods of wet and dry days showed the probability of the dry-days return period increases from the north to the south. An investigation of the role of general circulation in the creation of wet and dry periods showed that 4 patterns are effective in their production. Key-words: Markov chain, probability, dry and wet spells, Khuzestan Province.