A Precipitation-based Regionalization in Western Iran using Principal Component Analysis and Cluster Analysis

In the present study western Iran has been regionalized based on 10 factors in 170 stations using principal component analysis (PCA) and Cluster Analysis (CA). In this way, all 10 factors were reduced to 4 principal components and then rotated using Varimax rotation criterion. Applying Ward's Algorithm, a Hierarchical Cluster Analysis, on principal component scores, the stations were grouped into 5 individual clusters. The results indicated that the study area comprises of 5 distinctive homogenous subdivisions. Topography and latitude play an important role in determining boundaries between identified subdivisions and existence of spatial differences between them as well.