Temperature is a critical driver of ecosystem processes and food web dynamics. Past water temperatures may provide insights into mechanisms behind observed trends in aquatic ecosystems, and projections of future water temperatures can help prioritize management actions. Unfortunately, very few lakes are monitored for water temperature on the spatial scale and resolution necessary for these types of research questions. Air temperatures are widely available, but lake temperatures do not track air temperatures linearly. The decoupling of air and water temperatures is especially apparent in deep waters that provide critical habitat for cool- and cold-water fishes.

Together with collaborators at the United States Geological survey and elsewhere, we have developed a suite of tools for hindcasting and forecasting water temperatures in thousands of lakes. These include process-based models, deep-learning models, and hybrid approaches that combine the advantages of each.

Predicted lake temperatures are available for lakes in Minnesota, the Midwestern United States, and/or the coterminous United States.