Melting and Freezing Lake Ice: Getting Phase Transitions Correct in Lake Models

Transitional periods of phase changes on lake surfaces serve as important times of the year for heat exchanges between the boundary layer and the lake itself as the summertime energy partitioning becomes complicated by the presence of a dynamic ice sheet. These transitional periods have significance as a first step towards a lengthier period of total ice coverage on the lake surface, or no ice coverage on the lake surface; these two scenarios feature different interactions between the lake and the lower atmosphere, and thus require model adherence towards their respective relationships.

This project analyzes whether major lake models are able to accurately model interactions between the lakes and surrounding atmosphere, and thus if these models can be trusted during these critical periods of lake transition. This experiment uses two major lake models, GLM and FLake, in a comparison of modeled sensible and latent heat exchanges to observed data over a five-year period, using Lake Mendota in Madison, Wisconsin as the lake in question. The observational data has been collected by a flux tower stationed at the Center for Limnology on the southern shore of Lake Mendota, and will provide accurate heat flux data for comparison.

Good agreement between the models and the observational data has already been found during summer and wintertime periods due to a continuously present ice sheet or lack thereof, furthermore, year-round comparisons show good agreement as well. This leads the focus of this investigation to be smaller scale seasonal periods in fall and spring, during which the ice sheet is constantly changing. The experiment uses analytical and graphical comparisons to determine the precision of the models and observational data on both sub-daily and daily recordings, and ultimately determines if frequent phase changes on the lake surface can be accounted for by the two lake models. Comparing model output to observational data over a five-year window will provide insight into similar seasonal trends in addition to biases between different methods.

These transitional periods not only are important for scientific study, but they can be dangerous for society, as frequent phase changes between the water and surface ice sheet can be hazardous for people seeking to venture out onto the frozen lake surface, or people planning to go out boating unaware of ice formation. Our results illustrate the ability of the models to account for transitional periods and whether these models can be trusted as data for use in the scientific community, as well as for local weather reports regarding potential lake hazards.