The estimation of reference Evapotranspiration in Tabriz and Ardebil, by Principal Component Analysis (PCA)

Estimation of plant water requirement is one of the most important needs of the agricultural activity which can play an important role in the proper use of water resources. The first step for calculating of plant water requirement is the estimation of reference Evapotranspiration. According to this fact that the estimation of potential evapotranspiration needs lots of meteorological parameters, the aim of this research is to obtain a simple equation for estimating of evapotranspiration, using principal component analysis in Ardabil and Tabriz. For this aim parameters including temperature (maximum, mean and minimum), relative humidity, sunshine hours, precipitation and the wind speed in daily scale for a period of 1962-2016 for Tabriz and 1992-2015 for Ardebil is used.

The results of principal component analysis reduce these parameters to two and three components (PC) for Tabriz and Ardebil respectively. These PC explain the 78% of parameter’s variance in Tabriz and 83% in Ardebil, respectively. By using of these components, new equations are obtained for calculate the potential evapotranspiration. The results of evapotranspiration modeling show that the coefficient of determination between daily reference evapotranspiration and principal components (PC) for calibration and verification periods are 0.53 and 0.69 for Tabriz and 071 and 0.73 for Ardebil, respectively. Also, the Nash coefficients for Tabriz are 0.61 and 0.61 and for Ardebil are 071 and 0.73 which showing the appropriate performance of models. The results also show that the evapotranspiration in Tabriz is highly affected by temperature parameter, relative humidity and sunshine hours and in Ardebil is only affected by temperature.