Estimating Reference Evapotranspiration by Using Principal Component Analysis (PCA) and The Development of a Regression Model (MLR-PCA) (Case Study: Mashhad Station)

Evapotranspiration is one of the most important parameters that its understanding is necessary for estimating crop water requirement and design of irrigation systems. This phenomenon is greatly influenced by climatic parameters. In this study, the relative importance of variables affecting this phenomenon was evaluated and the reference evapotranspiration was estimated using principal component analysis and factor analysis. Daily scaled measurements for the period of 1991-2005 were obtained from synoptic stations located in Mashhad Khorasan Razavi provience, Iran. Mashhad has a semi-arid climate area. The measurements included the relative influence of temperature (T) (maximum, average and minimum), relative humidity (RH), sunshine hours (Rs), and the wind speed at a height of two meters above the ground (U2). The multiple linear regressions were used to estimate evapotranspiration. T-statistic with a significant level of 5% was used for the main components. The evapotranspiration was correlated more with T (minimum. maximum, and average), and relative humidity as than wind speed or sunshine. PC1 had more effect than PC2 (with coefficients of 0.694 and 0.556, respectively). MLR-PCA and MLR with coefficients of 0.903 and 0.897 (respectively) indicated higher ability for PCA method.