Estimation of Monthly Mean Maximum and Minimum Soil Temperatures from Profiles Taken at 9:00 A.M.

Abstract

At many agricultural meteorological stations, soil temperatures are observed once a day at 9:00 A.M. In the upper soil layers, however, the temperature fluctuates widely during the course of a day, and the vertical profile taken at a time of day is not sufficient for an understanding of the soil temperature regime. Thus, it is desired to evolve a method for utilizing the observations made at 9:00 A.M. as fully as possible.We constructed a simple numerical method for the purpose of estimating the monthly mean maximum and minimum temperatures in the upper soil layers from the monthly mean profile taken at 9:00 A.M., and examined the accuracy of the estimation, using the data obtained in the bare observation fields of Shimane University (clay) and Sand Dune Research Institute, Tottori University (sand).In the present method, monthly mean curves of diurnal temperature variations at all depths are approximated by pure sinusoidal functions of time around a mean which is constant throughout the layers. The approximation becomes better with increasing depth. The function has four parameters, that is, the daily mean, the amplitude at the surface, the damping depth and the instant at which the surface temperature is at its mean and increasing. The last one averaged over a month can be estimated within half an hour from the time of sunrise; the instant comes about 2 hours and 15 minutes after sunrise in bare clay fields and 15 to 30 minutes further behind it in bare sand fields. The other three parameters are estimated numerically from the instant and the monthly mean temperatures observed at three depths at 9:00 A.M.; then the maximum and minimum temperatures are calculated by using the estimates of the parameters.Results obtained by the analyses of the two sets of data above are similar in trend. The estimated values of monthly mean maximum and minimum temperatures at a depth of 5cm are about 1 to 3°C lower than the observed ones in the spring and summer months, but are in fair agreement with the observations in the autumn months. The accuracy of the estimated values at 20cm depth is about ±1°C for the maximum and about ±0.3°C for the minimum. The accuracy is quite satisfactory in view of the approximated character of the present method and the fact that only the time of sunrise and the soil temperature profile taken at 9:00 A.M. are needed to utilize the method.