A simple model for predicting daily mean soil temperatures

Berlin

1989

312-318

Sonderdruck

12509S

203891

Based on measurements in 3 different types of soil (clay, sand, peat) linear regression equations between daily air temperature (2 m) and soil temperature (2, 5, 10, 20, 50 cm depth) are calculated for all months of the growing season. The equations show a significant seasonal dependence and the best correlations in the upper 10 cm of soil. Differences depending on the type of soil are relatively small. Correction terms involving cloudiness and thermal inertia of the soil during a sudden warming or cooling period complete the prediction model. Standard deviations between predicted and measured values have been found within 1.5 K in most cases. Lastly a generally applicable method for calculating regression equations at any station is introduced. The application of this method to different sites and types of soil in Bavaria and other regions of Germany shows a good agreement with measured values.