Computational Models for Average Air and Soil Temperature Predictions

Abstract

Long term data on soil temperature (ST) are most often not available and can be predicted from air temperature data for use in crop planning. Daily soil temperature data of morning and evening for three years (1970 -1972) and daily maximum and minimum temperature data for 20 years (1967 -1986) of Jabalpur were collected. Models were developed for computation of average Soil temperature (ST). ST during year using the daily morning and the daily evening ST of ith day of week over the period considered. The models were also developed for computation of the average air temperature of ith week using daily minimum and maximum temperature. Regression models were developed for estimation of average weekly ST from maximum, minimum, average air temperature or the parameters derived from air temperature. Mostly, average air temperature is used in predicting ST but the present study indicated that only average weekly maximum temperature could predict ST for all the weeks with greater accuracy. Thus, a single equation can predict ST for all the weeks when the average weekly maximum temperature is used as input parameter.