عنوان مقاله [English]
Study of humidity is important from different views such as agricultural productions, architecture, human welfare, water resources studies and rainfall. This study focuses on humidity Using Multi-Variable Statistical Methods (principal component analysis and cluster analysis) in the North West of Iran. Therefore, seasonal quantities of ten climate variables in 20 synoptic stations in North West of Iran during 20 years were studied. These variables include mean, minimum and maximum relative humidity, relative humidity in 3, 9 and 15 UTC, water vapor pressure, mixing ratio, saturation deficit and dew point). Thus the 40*20 matrix was formed. Statistical data were normalized and due to different scales of data, the standard scores were used in analysis. The results of principal component analysis showed that 4 components represent about 96% of the original variables variance. After identification of these four components, the spatial pattern of Component scores was plotted. Analysis of this maps showed that the spatial and temporal distribution of moisture in this area are not the same. Data component matrix scores used hierarchical method for clustering and five humidity zones were separated and mapped.