Recognition of homogeneous regions of heavy and super heavy precipitation in Iran by intergroup variance quality control indices

The aim of this study is spatial clustering of the rainfall patterns of Iran. In order to perfrom this research daily interpolated precipitation from Asfazari** database during 21/03/1961 to 31/12/2004 has been used. Considering the country’s area and study period, a database with dimension 159927187 has been created with time (days) as the rows and spatial location as the columns’ pixels. For every calendar day of year and each pixel, heavy and super heavy precipitation have been calculated based on 95 and 99 percentiles. Besides, two data base with dimension 7187366 obtained. Regionalization was performed using cluster analysis with Ward linkage method. For determining number of clusters, 11variance inter groups quality control indices have been applied on the different clustering. The results showed that Iran can be clustered to five homogeneous regions of heavy and super heavy precipitation. Daily and monthly precipitation thresholds were also calculated for each region. In general, the results indicated that applied quality control indices have are suitable for spatial analysis and detecting number of homogeneous regions of heavy and super heavy precipitation across Iran.