Main content area

Modeling soil temperature in a temperate region: A comparison between empirical and physically based methods in SWAT

Qi, Junyu, Zhang, Xuesong, Cosh, Michael H.
Ecological engineering 2019 v.129 pp. 134-143
Soil and Water Assessment Tool model, agricultural watersheds, heat transfer, model validation, monitoring, soil depth, soil surface layers, soil temperature, temperate zones, winter, Choptank River, Maryland
Although the Soil and Water Assessment Tool (SWAT) has been widely used in temperate regions, the performance of its soil temperature module has not been extensively assessed. The aim of the present study is to evaluate the performance of the SWAT model’s built-in empirical soil temperature module and a physically-based soil temperature module using four years of daily soil temperature measurements at three depths (i.e., 5, 10, and 50 cm) across 10 monitoring stations in and around the Choptank River Watershed, Maryland, USA. Model performance is assessed according to three coefficients of accuracy, i.e., Bias, Nash-Sutcliffe coefficient (NS), and coefficient of determination (R2). Model performance of SWAT and the modified version of SWAT (i.e., MSWAT; equipped with the physically-based soil temperature module) is also compared in winter and non-winter seasons. Results show that R2 and NS for different soil depths across 10 stations are all greater than 0.95 and 0.90, respectively, for both SWAT and MSWAT. This indicates that both SWAT and MSWAT reproduce well variations of measured soil temperatures at all soil depths for all stations. The results also show that SWAT and MSWAT tend to underestimate soil temperatures in both winter and non-winter seasons at all soil depths across 10 stations, and MSWAT improves soil temperature simulation by reducing absolute values of Bias, especially at surface soil layers in winter. We found out that a better representation of surface residue is needed in the physically-based soil temperature module for applications in agricultural watersheds of temperature regions. Without a physical accounting, a calibration procedure may be required to account for residue effects on heat transfer based on the current version of the physically-based module.