Validation of a 1-D lake model for modeling evaporation from an elongated and deep boreal reservoir

Freshwater reservoirs modify the regional climate through mass, energy, and momentum exchanges with the atmosphere. Recent studies have shown that hydropower reservoirs tend to evaporate more than the land they have flooded, hence reducing water availability for other uses, at a level that should with local climate conditions, however. Knowing that evaporation is a key component of the water balance and that very few studies have focused on evaporation from northern reservoirs, which are ice covered several months per year, there is a real need for models that can provide reliable estimates of this water vapor flux. This project focuses on the modeling of evaporation from an 85-km² hydropower reservoir located in the boreal biome of eastern Canada (50.7°N, 63.2°W), with a mean depth of 60 m and an elongated shape. To support this modeling effort, two flux towers (one on the shore and one on a raft) and a vertical chain of thermistors were deployed. Exchanges between the water surface and the atmosphere are simulated with the Canadian Small Lake Model (CSLM), a 1-D physical-based surface scheme designed to be coupled with a numerical weather prediction model. The model also simulates the thermal regime of the water body, including ice formation. Considering the irregular shape of the reservoir as well as its depth, a new model parameterization was adopted that improved simulations (albedo parameterization, leakage parameter, mixed layer maximum depth...). Turbulent fluxes were successfully predicted during the open water period. Comparison between observed and modeled time series showed a good agreement specifically for sensible heat fluxes.  Deviations mostly occur before freeze-up (October to November) and around ice off (April to May) with a tendency of overestimating latent heat fluxes when its observed magnitude is small (ice period). Thermal mixing as well as mixed layer deepening were well estimated. Thermal mixing, as well as mixed layer deepening, were well estimated. Near-surface water temperature confirmed the ability of the CSLM to simulate the near-surface seasonal cycle. However, in early fall, an overestimation of the water temperature induced an overestimation of the heat fluxes leading to early depletion of the energy storage that led to an early modeled freeze-up.