ESTIMATING SOIL SURFACE TEMPERATURE FROM METEOROLOGICAL DATA

Abstract

Models that utilize meteorological data to calculate temperature near the ground surface are useful to implement historic weather records into current biological or soil model development. The objective of this research was to develop and validate a method for supplementing existing experimental data sets with surface temperature. A surface energy balance approach together with an empirical function relating surface resistance to soil wetness was used to predict surface temperature. Model inputs are relative humidity, net radiation, accumulated net radiation since the last rain, air temperature, and wind speed. Predicted surface temperatures at two sites were compared to soil temperature measured at 2.5 mm and surface temperature measured with an infrared thermometer. Predicted soil water evaporation during late summer was compared to measured soil water evaporation when the surface to 9-cm soil layer was dry (0.06 cm3/cm3) and wet (0.15 cm3/cm3). Standard deviations of the hourly differences between predicted and measured surface temperature were less than 2.5°C for both bare soil and residue cover treatments. Standard deviation of the difference between measured and predicted evaporation was 0.15 mm/day. This model gave a reasonably good estimate of ground surface temperature. Results also suggest that the model correctly partitions net radiation to sensible, soil, and latent heat flux.