MACROCLIMATIC MODEL FOR ESTIMATING MONTHLY SOIL TEMPERATURES UNDER SHORT-GRASS COVER IN CANADA

Abstract

A macroclimatic model was developed to estimate monthly soil temperatures under short-grass cover. It involved multiple regression equations for each month and for each of six depths (1, 10, 20, 50, 100, and 150 cm). Data used were obtained from published records of soil temperature and corresponding climatic variables. They were from 41 stations over several years with station-years per regression varying from 88 to 226 according to depths and months. The climatic variables were related to air temperature, rainfall, snowfall, and potential evapotranspiration. An additional important variable was the estimated soil temperature of the previous month. The equations explain 70–96% of the soil temperature variations and the standard errors of estimate varied from 0.7 to 2.2 C. Temperatures estimated for 1 yr and eight stations with climatic data not used in the development of the equations departed from the observed values by less than 0.5, 1.0, and 2.0 C in 34, 62, and 92% of the cases, respectively. Errors resulting from the estimation of monthly normals by this model are expected to be generally less than 1.0 degree C.