Testing and comparison of three unsaturated soil water flow models

Abstract

The performance of three soil water flow models is compared using field data collected from a structureless sandy loam soil cropped to soybeans (located near Simcoe, Ontario, Canada), and a structured clay soil cropped to grass hay (near Ottawa, Ontario, Canada). Two of the models, Soil Water and Actual Transpiration Rate, Extended (SWATRE) and Leaching Estimation and Chemistry Model (LEACHW), are based on the one-dimensional pressure head form of Richards' equation, and the third, SWASIM, is based on the one-dimensional diffusivity-water content form of Richard's equation. The models are tested and compared on the basis of their ability to predict in situ measured soil water content profiles during the growing season at the two field sites. All three models produce acceptable predictions, as evidenced by an average error within ±0.0404 cm 3·cm −3 and a relative error within ±0.2915. None of the models is consistently more accurate than the others. Discrepancies between the models are attributed primarily to differences in how the models account for soil evaporation and plant transpiration.