Understanding drivers of recent record-high water levels across the Laurentian Great Lakes

Over the last two decades, water levels of the Laurentian Great Lakes have exhibited an accelerated transition from extreme lows to extreme highs, posing a challenge to regional water resources management. Improved understanding of the factors driving changes in water levels of the Great Lakes is essential to enable the modification and advancement of new water policies in the context of climate change. This study investigated the variation of water budget components and their contributions to changes in water levels across the Great Lakes, with a focus on understanding changes in the peak water level season. To identify drivers of recent water level changes, we developed a new seventy-year long record of monthly water budget data using the Large Lakes Statistical Water Balance Model, a recently developed model encoded in a Bayesian Markov chain Monte Carlo framework. The recent surge in water levels was observed consistently over the Great Lakes system, but the key mechanism driving this pattern varies substantially across individual lakes. The new water budget estimate, with the capacity to reconcile the water balance, highlights the benefit of moving to a probabilistic framework in water resources accounting for the Great Lakes. The Large Lakes Statistical Water Balance Model, considering its capacity to accommodate multiple estimates of each water balance component, provides a mean for operational agencies in the U.S. and Canada to resolve discrepancies in hydro-meteorological data for the Great Lakes. The presented framework can also serve as a useful asset to benchmark new estimates of the Great Lakes water balance components such as over lake precipitation or evaporation.