Enhancing Ecosystem Modeling in Lake Erie through Improved Thermal Structure Simulation via Data Assimilation

Water temperature is a fundamental characteristic of lake ecosystems, influencing all physical, chemical, and biological processes. To comprehensively understand and accurately predict a lake's biophysical properties, it is essential to study the spatial and temporal variations in its thermal structure and surface temperature. This study focuses on Lake Erie, aiming to assess if and to what extent improvements in thermal structure simulation through data assimilation (DA) can enhance the performance of ecosystem modeling. A 3-D hydrodynamic model (FVCOM) was employed to simulate the physical processes in Lake Erie, providing essential hydrodynamic data—such as current velocity, depth, and water temperature—to an ecosystem model. The optimal interpolation (OI) method, one of the most efficient DA techniques, was applied to the hydrodynamic model to improve both hindcasts and forecasts of water temperature. The ecosystem model used in this study was based on the Lake Erie Ecosystem Model (LEEM), and made some improvements, including but not limited to refined relationships between temperature/light and growth rates, a critical factor in this study. Through several numerical experiments, this research aims to demonstrate that accurate temperature modeling is foundational to precise ecosystem modeling, with water temperature recognized as a key input parameter.