This study introduces HyLake, a hybrid lake model that embeds a deep-learning surrogate for the water temperature module within a process-based backbone. HyLake simulates lake surface temperature and the latent and sensible heat fluxes in Lake Taihu more accurately than traditional process-based models and other hybrid experiments across different forcing datasets. The proposed coupling strategy provides a reliable tool for quantifying the impacts of climate change on aquatic ecosystems.

Regional analyses of atmospheric trace gases can improve knowledge of fluxes at high resolution but rely on specified boundary conditions (BCs) at the domain edges. Biases in the often-uncertain BCs propagate to the inferred fluxes. We develop a framework to explain how errors in the BCs influence the optimized fluxes, derive two metrics to estimate this influence, and compare two methods to correct for the biases. We demonstrate correcting BCs directly is more effective at reducing bias.