عنوان مقاله [English]
Temperature is an important soil physical parameter and could control soil physical and chemical processes as well as the growth and yield of crops. The aim of the present study was to use the sinusoidal model for predicting soil temperature in four selected synoptic stations with different climate, namely, Bandar Abbas, Rasht, Shahrekord, and Yazd stations. Soil surface (5 cm) daily temperature was collected during ten years period i.e. 1998 to 2007.For modeling, data set was divided into two parts: 70% for training and 30% forvalidation. To eliminate the effects of local climate variations on modeling process, data set of 1998 to 2002, 2004 to 2005 and 2007 were chosen for modeling. Daily data set of 2000, 2003 and 2006 were used for model evaluation. Two input parameters of sinusoidal model including mean and annual ranges of temperature were calculated using arithmetic mean, minimum, and maximum temperatures of the data set.Validation indices including RMSE, MBE, MPE, d, and R2 calculated for the four stations were, on average, 2.64, -0.16, 10.84, 0.59, and 0.93, respectively. The result of sinusoidal modeling showed that the model efficiency decreased for Bandar Abbas, Yazd, Shahrekord, and Rasht stations, respectively.