Spatial Autocorrelation Analysis of Extreme Precipitation in Iran

M. Darand, M. Dostkamyan and M. I. A. Rehmani

Spatial variations in extreme precipitation events make hydrological, climatological, social, environmental and agricultural effects on a country. This study presents the spatiotemporal autocorrela- tion analysis of extreme precipitation events over Iran using gridded data on daily precipitation for the period 1961–2010. The 95th percentile is considered as extreme precipitation factor. The spatial auto- correlation of extreme precipitation is examined by three commonly used spatial autocorrelation statistics, the Gi statistic index, Moran’s I global index, and Local Moran’s I (LISA) index, at the 95 and 99% significant confidence level. The results showed a strong significant spatial autocorrelation for extreme precipitation events with the highest Moran’s I value in January. The positive significant autocorre- lation of extreme precipitation is observed over the southern parts of the Caspian Sea and Zagros Mountains ranges, while the negative significant autocorrelation is observed over the central and eastern parts of country. In spring and summer the positive autocorrelation cores displace from the Zagros Mountains ranges to the northwestern and southeastern parts.