Search

Abstract Limnological understanding of the role snow plays in under‐ice thermal dynamics is mainly based on studies of clear‐water lakes. Very little is known about the role snow plays in the thermal dynamics of dystrophic lakes. We conducted a whole lake experiment on a small, 8 m deep dystrophic bog lake in northern Wisconsin, where we removed all snowfall over two consecutive winters. Due to weather variability, only 1 year had predominantly black ice. Under these conditions, the lake rapidly cooled in early and mid‐winter, compared to snow covered conditions that insulated the lake from heat loss. The lake also rapidly gained heat in late winter resulting in isothermal conditions well in advance of ice‐off. These results show how water clarity modulates the influence of snow on under‐ice thermal dynamics, which is relevant to futures with snow droughts.

Abstract Hybrid Knowledge‐Guided Machine Learning (KGML) models, which are deep learning models that utilize scientific theory and process‐based model simulations, have shown improved performance over their process‐based counterparts for the simulation of water temperature and hydrodynamics. We highlight the modular compositional learning (MCL) methodology as a novel design choice for the development of hybrid KGML models in which the model is decomposed into modular sub‐components that can be process‐based models and/or deep learning models. We develop a hybrid MCL model that integrates a deep learning model into a modularized, process‐based model. To achieve this, we first train individual deep learning models with the output of the process‐based models. In a second step, we fine‐tune one deep learning model with observed field data. In this study, we replaced process‐based calculations of vertical diffusive transport with deep learning. Finally, this fine‐tuned deep learning model is integrated into the process‐based model, creating the hybrid MCL model with improved overall projections for water temperature dynamics compared to the original process‐based model. We further compare the performance of the hybrid MCL model with the process‐based model and two alternative deep learning models and highlight how the hybrid MCL model has the best performance for projecting water temperature, Schmidt stability, buoyancy frequency, and depths of different isotherms. Modular compositional learning can be applied to existing modularized, process‐based model structures to make the projections more robust and improve model performance by letting deep learning estimate uncertain process calculations.