Estimation of Some Climatic Parameters by Using Pedo-Transfer Functions

Some Iranian weather stations do not have sufficient instruments to measure climatic parameters such as pan evaporation, solar radiation and sunshine hours and/ or lack long-term and continuous data. A Pedo-transfer function (PTF) is an indirect method to predict parameters that are difficult to measure. Multiple Linear Regression (MLR) and Artificial Neural Network (ANN) were used to develop the PTFs. In this study, application of PTFs for estimation of climatic parameters was conducted for synoptic weather station of Tabriz. The objective of this study was to predict monthly parameters as solar radiation, evaporation, and sunshine hours from readily available parameters of weather. The results indicated that PTFs estimated solar radiation, evaporation, and sunshine hours well. The performance of both MLR and ANN were almost similar, only for solar radiation, ANN predictions were better than regression equation. The estimation of pan evaporation and sunshine hours were better than solar radiation in both PTF functions (MLR and ANN).