Wind Correction in 3D Modeling of Small Lakes with Tree Canopies

Abstract Modeling effects of surrounding tree canopy and bluff topography on the wind introduction onto lakes surface remains an ongoing challenge in three-dimensional (3D) hydrodynamic modeling and requires extra costly modeling (e.g., via CFD) prior to the main modeling. We studied and suggested different less costly approaches in such lakes by using the 3D aquatic ecosystem hydrodynamic model. We modeled two adjacent groundwater-fed lakes, Sacrower See (SCR) and Gross-Glienicker See (GGS) in northeast Germany, over a decade. The two lakes with different morphologies are located within a similar forest, climatic, and hydrologic system. The effects of calibrating the wind drag coefficient ( C D ), validity of the Markfort correction, originally developed for one-dimensional (1D) models, and modifying the fetch on the model grid were evaluated and compared to the uncalibrated model and measured water temperature profiles and further compared to an equivalent round lake mimicking quasi-1D conditions in a 3D model. The results show that the Markfort approach does not change or may slightly increase the accuracy of 3D models in the case of the deeper lake (SCR). However, it significantly reduces the accuracy of the model in a shallow lake (GGS) due to neglection of the wind direction and higher sensitivity of the shallower lake to wind. Our results indicate that calibrating C D in tandem with calibrating the heat transfer coefficient ( C H ) results in the highest accuracy in a 3D model of a lake with tree canopy and may be considered a less time- and power-consuming approach, as these two coefficients reflect the most important surface exchanges between the lake and the atmosphere. By modeling two lakes with different morphologies and slightly different surroundings, we showed that our findings are applicable to different lakes.