Understanding the performance of three 1-D lake models in simulating thermal dynamics of diverse water bodies in the Yangtze River Basin

Abstract

Understanding the thermal processes of lakes and reservoirs is essential due to their profound influence on local climate and environment, as well as their sensitivity to climate change and human activities. This study aims to systematically compare the performance of three one-dimensional lake models (FLake, CoLM-Lake, and Simstrat) in simulating the thermal dynamics of different water bodies in the Yangtze River Basin, which contains the highest water bodies density in China. The results showed that while all models performed well for shallow water bodies, significant challenges and variability emerged for deep water bodies, with Simstrat outperforming CoLM-Lake and FLake in capturing thermal structure. Specifically, Simstrat demonstrated superior performance in simulating the vertical layering of water caused by differences in temperature and density (i.e., thermal stratification) in Lake Qiandaohu (RMSE: 1.44 °C) and surface temperatures across deep water bodies (average RMSE: 2.16 °C). CoLM-Lake effectively reproduced surface temperatures in deep water bodies (average RMSE: 2.36 °C) but overestimated vertical mixing, leading to less accurate stratification simulations (RMSE: 3.16 °C). FLake exhibited limitations in simulating temperatures and thermal stratification in deep water bodies but performed relatively well in shallow systems. Moreover, all three models exhibited diminished accuracy in reservoirs simulations compared to lakes, possibly due to inadequate representation of key processes. Additionally, we explored the impacts of different surface flux schemes and parameter calibration strategies on model performance. This study offers crucial insights into enhancing the simulation of thermal processes in lakes and reservoirs, particularly for deep-water environments, thereby advancing our understanding of thermal dynamics and their implications across the Yangtze River Basin.