A Thermally Stratified Reservoir Module for Large‐Scale Distributed Stream Temperature Models With Application in the Tennessee River Basin. Issue 10 (19th October 2018)

• Record Type: Journal Article
• Authors: Niemeyer, Ryan J.; Cheng, Yifan; Mao, Yixin; Yearsley, John R.; Nijssen, Bart
• Is Part Of: Water resources research. Volume 54:Issue 10(2018)
• Page Start: 8103
• Page End: 8119
• Publication Date: 2018-10-19
• Subjects: stream temperature -- RBM -- thermal stratification -- reservoir -- hydrologic modeling; Hydrology -- Periodicals; 333.91
• DOI: 10.1029/2018WR022615
• Languages: English
• ISSNs: 0043-1397

Abstract

Abstract: River temperatures affect water quality, power plant cooling, and freshwater ecosystems. Stream temperature models that treat river reaches and reservoirs as well‐mixed segments do not capture thermal stratification in reservoirs. To account for the effects of reservoir stratification on downstream water temperatures, we developed a two‐layer stratified reservoir module, which was integrated into the River Basin Model (RBM) to simulate river temperature across a river network with multiple large thermally stratified reservoirs. To evaluate the performance of this model configuration compared to RBM without thermally stratified reservoirs, we simulated river temperature in the Tennessee River Basin in the southeastern United States. We simulated land surface hydrologic fluxes using the Variable Infiltration Capacity (VIC) model and routed runoff using the river routing model RVIC. The two‐layer model configuration reduced the bias in simulated summer river temperature from 6.7 to −1.2 °C downstream of a reservoir with a residence time of 92 days and from 3.0 to −0.7 °C downstream of a reservoir with a residence time of 8 days. Improvement in fall and winter, when reservoirs tend to be well mixed, is minimal. RBM with the two‐layer module also captured the observed longitudinal river temperature variation downstream of a reservoir, with cool temperatures immediately downstream of the reservoir and gradual warming of the river as it flows downstream. Incorporating a simple stratified reservoir module into RBM improves model performance and increases the ability to apply the river temperature model to large basins with multiple large reservoirs. Plain Language Summary: River temperature is important for ecosystem health and water quality. It is affected by natural processes and also by human interventions, such as man‐made infrastructure. In particular, the construction of dams has established large and deep reservoirs where the temperature near the surface is warmer during summer than near the bottom. This change of temperature with depth is a natural process, called thermal stratification. Since water is typically released from the deeper reservoir layers, the water temperature downstream of these dams is often cooler during summer than it would have been without dams. This stratification process is well known and has been modeled before, but we are interested in capturing this process in large river systems with many large dams. In this paper, we develop a new component to represent the effects of stratification on reservoir and stream temperature and add it to an existing computer model that simulates stream temperature. We simplify the problem by representing the reservoir as a two‐layer system and demonstrate that the model matches the behavior of observed stream temperatures in the Tennessee River Basin. The resulting model is suitable for application over river basins with many reservoirs. Key Points: A stream temperature model coupled with a two‐layer reservoir module can efficiently simulate temperature for a multireservoir system River temperature simulations are generally improved, especially downstream of reservoirs with longer residence times