Comparison of Performance of Two Simulation and Regression Models for an Estimation of Soil Temperature under Grass Cover in Karaj Climatic Conditions

Because of the scarcity of in situ measurements، estimation of soil temperature by other means is very indispensable، as for Irrigation management and scheduling when in different field conditions. So far، many regression models have been developed for an estimation of soil temperature، using meteorological data under bare soil. Throughout this study، the performance of COUP simulation model and multiple regression approach for an estimation of soil temperature within an experimental plot، and under grass (Lolium perenne) canopy (in Karaj climatic conditions has been evaluated. Soil physical parameters، estimated as based on soil analysis (soil texture، bulk density)، and daily meteorological data (including maximum and minimum temperature، wind speed، pan evaporation، sunshine hours and rainfall) as well as vegetation data (crop height، root depth and Leaf Area Index (LAI) were made use of to run the model over the growing period. Soil temperature was measured using standard soil thermometers at depths of 10، 30، 50 and 70 centimeters. Stepwise approach was employed to develop suitable regression models. Following a running of both simulation and statistical models، the observed and simulated data values were compared، making use of statistical indices. The results revealed that، by inclusion of variables affecting incoming radiation i. e. crop height، and leaf area index، the accuracy in the prediction of soil moisture increases.