Location: Northwest Watershed Research Center
Title: Evaluation of 14 frozen soil thermal conductivity models with observations and SHAW model simulationsAuthor
HE, HAILONG - NORTHWEST A&F UNIVERSITY | |
Flerchinger, Gerald | |
KOJIMA, YUKI - GIFU UNIVERSITY | |
HE, DONG - NORTHWEST A&F UNIVERSITY | |
Hardegree, Stuart | |
DYCK, MILES - UNIVERSITY OF ALBERTA | |
HORTON, ROBERT - IOWA STATE UNIVERSITY | |
WU, QINGBAI - CHINESE ACADEMY OF SCIENCES | |
SI, BINGCHENG - UNIVERSITY OF SASKATCHEWAN | |
LV, JIALONG - NORTHWEST A&F UNIVERSITY |
Submitted to: Geoderma
Publication Type: Peer Reviewed Journal Publication Acceptance Date: 5/7/2021 Publication Date: 5/16/2021 Citation: He, H., Flerchinger, G.N., Kojima, Y., He, D., Hardegree, S.P., Dyck, M., Horton, R., Wu, Q., Si, B., Lv, J. 2021. Evaluation of 14 frozen soil thermal conductivity models with observations and SHAW model simulations. Geoderma. 403. Article 115207. https://doi.org/10.1016/j.geoderma.2021.115207. DOI: https://doi.org/10.1016/j.geoderma.2021.115207 Interpretive Summary: Soil thermal conductivity is a critical thermophysical property that is required for a variety of science and engineering applications. Numerous algorithms exist to estimate thermal conductivity of frozen soils, but no known study has been conducted that incorporates multiple thermal conductivity algorithms into a numerical simulation model for direct comparison of simulation results with measured field data. In this study, 14 thermal conductivity algorithms were implemented into the Simultaneous Heat and Water (SHAW) model and simulated soil temperatures were compared with field observations at two sites with very different soil texture (i.e., sand, silt and clay). Errors ranged from -2.2 to 2.8 ºC, which was not deemed very satisfactory. Future study is required to develop new frozen soil thermal conductivity algorithms for more accurate and wider range of applications. Technical Abstract: Frozen soil thermal conductivity (FSTC, 'eff) is a critical thermophysical property that is required for a variety of science and engineering applications. Measurement of 'eff in frozen soils is prone to errors, especially near the freezing point of soil water (e.g., -4 to 0 ºC) and few data are available. Therefore, many FSTC algorithms have been developed and a few of them have been incorporated in numerical simulation programs to better understand energy balance at the ground surface or soil thermal regime. However, large discrepancies between simulated and observed values have been reported. Previous studies either evaluated the performance of a few FSTC algorithms with a small experimental dataset of 'eff or simply compared their performance in the numerical simulation programs. No study has been conducted to systematically assess the performance of the FSTC algorithms included in numerical simulation programs with both experimental datasets and model simulations. Our extensive literature review returned 14 FSTC algorithms that have been incorporated in a variety of numerical simulation programs. The 14 FSTC algorithms were evaluated with a compiled dataset consisting of a total of 331 measurements at temperatures below -4 ºC on 27 soils from seven studies. The results showed that the 1992 Becker algorithm performed best on the experimental dataset but was still not satisfactory (RMSE=0.46 W m-1 ºC-1, Bias = -0.04 W m-1 ºC-1 and NSE=0.51). The inclusion of liquid water content could slightly increase the performance of the FSTC algorithms rather than assume all water becomes ice. These FSTC algorithms were also incorporated in the SHAW model to compare their effects on the simulation of soil temperature and water content on two sites with contrasting soil textures. The simulation results showed that the average bias of simulated and observed soil temperature for all depths ranges from -2.2 to 2.8 ºC and the average differences of liquid water content range from -0.08 to 0.1 cm3 cm-3. Generally no FSTC algorithm could be used to satisfactorily simulate the soil thermal regime. Future study is required to develop new FSTC algorithms for more accurate and wider applications. |