Coupling terrestrial and aquatic thermal processes for improving stream temperature modeling at watershed scale

Authors: QI, Y. - University Of Maryland; LEE, S. - University Of Seoul; DU, X. - University Of Alberta; FICKLIN, D.L. - Indiana University; WANG, Q. - Global Change Research Institute; MYERS, D. - Indiana University; SINGH, D. - Oak Ridge National Laboratory; Moglen, Glenn; McCarty, Gregory; ZHOU, Y. - Iowa State University; Zhang, Xuesong

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/19/2021
Publication Date: 9/28/2021
Citation: Qi, Y., Lee, S., Du, X., Ficklin, D., Wang, Q., Myers, D., Singh, D., Moglen, G.E., Mccarty, G.W., Zhou, Y., Zhang, X. 2021. Coupling terrestrial and aquatic thermal processes for improving stream temperature modeling at watershed scale. Journal of Hydrology. 603. Article 126983. https://doi.org/10.1016/j.jhydrol.2021.126983.
DOI: https://doi.org/10.1016/j.jhydrol.2021.126983

Interpretive Summary: Water temperature in streams and rivers is an important water quality indicator, which can influence the rates of chemical reactions, metabolic rates, and biological activity of aquatic organisms. The USDA Soil and Water Assessment Tool (SWAT) model uses simple empirical relationships to represent thermal processes on land and in water. We revised the SWAT model to include new modules that are capable of explicitly depicting physical processes in terrestrial and aquatic ecosystems to determine stream water temperature. Using observed stream temperature data from 12 sites, we tested the new method and compared its performance with previous versions of the SWAT model in watersheds in the US mid-Atlantic and in the US Midwest. Findings of this study suggest that the new method not only performs equal or better than algorithms used in previous studies, but also is capable of better representing processes related to how climate change affects stream temperature. The revised SWAT model will be useful to inform ecosystem services for watershed assessment and planning.

Technical Abstract: Water temperature is an important indicator of the health of terrestrial and aquatic ecosystems and influences numerous biological and chemical processes that affect water quality, the health and productivity of biota, and other ecosystem services. In this study, we improved the Soil and Water Assessment Tool (SWAT) model to explicitly consider the thermal processes influencing the temperature of various terrestrial water components (i.e., surface runoff, lateral flow, and base flow) and the heat balance of water in streams. The improved SWAT model with an explicit Heat Balance description on both terrestrial and aquatic thermal processes (SWAT-HB) was compared with previous versions of the SWAT model that employ only Empirical Relationships between air temperature and stream temperature (SWAT-ER) or combine Empirical terrestrial water temperature estimates with the Heat balance of stream water (SWAT-EH) for the Upper Mississippi River Basin (UMRB) and Greensboro Watershed (GW) in the United States. The results show that, even though the SWAT-HB better simulates terrestrial thermal processes, all three models produce comparable simulations of long-term stream temperatures, indicating that stream temperature is mainly influenced by air temperature as water travels along large rivers. Nevertheless, the SWAT-HB and SWAT-EH models achieved better performance than the SWAT-ER model in the winter and spring but exhibited mixed performances in the summer and fall. The better model performance for the SWAT-HB model than the SWAT-EH model can be attributed to the improved estimation of terrestrial water temperature components. Although, empirical methods (e.g., SWAT-ER) are suitable for providing estimation of stream temperature, the coupled terrestrial and aquatic heat balance-based methods (e.g., SWAT-HB) allow for the consideration of land and water management practices (e.g. crop irrigation, land use change, and water use for thermoelectric cooling) effects on stream temperature.