<p>Observations of sensible and latent heat fluxes over inland water bodies are unfortunately scarce and, yet, critical to the development of adequate lake parameterization for numerical weather forecast and climate models. When available, they usually consist of eddy covariance (EC) or lysimeter measurements, both representative of a relatively small footprint area, typically of a few hectares in the case of the EC approach. Over the past decades, we have seen the emergence of bichromatic scintillometry (SC), which allows for a &#8216;regional&#8217; (~km<sup>2</sup>) estimation of turbulent heat fluxes. In brief, two beams travelling from a set of transmitters to a set of receivers scintillate in the turbulent air above the surface of interest and enable, using the Monin-Obukhov Similarity Theory, the computation of sensible and latent heat fluxes at the land-atmosphere interface. While a handful of studies have looked at the performance of this approach over land surfaces, very few have assessed it over water bodies. This study presents an evaluation of scintillometry-derived turbulent heat fluxes over an 85-km<sup>2</sup> boreal hydropower reservoir of eastern Canada (50.69&#176;N, 63.24&#176;W) with respect to those obtained with EC measurements collected on a nearby floating platform. The scintillometer beam path travelled for 1.7 km over a surface of the reservoir that reached depths of ~100m, from 14 August to 9 October 2019. Results indicate positive, day-and-night, latent heat fluxes throughout the whole period; highlighting that the reservoir steadily released heat over the second half of the open water period, from mid-august until freeze-up. Sensible heat fluxes peaked at night due to the near-surface air temperature vertical gradient reaching its daily maximum. For sensible heat fluxes, the SC approach corroborates well with the EC approach, while for latent heat fluxes, the agreement between EC and SC decreases. This suggests that the larger footprint of the SC system might be affected by heterogeneous surface flux characteristics in the reservoir, which encapsulates the need for large-scale measurements. Grouping results by atmospheric stability regimes does not improve comparison results. These results provide an opportunity to validate an innovative approach for measuring turbulent fluxes at a regional scale and, hence, improving our understanding of turbulent fluxes over large reservoirs and lakes.</p>
Feasibility of Estimating Turbulent Heat Fluxes via Variational Assimilation of Reference-Level Air Temperature and Specific Humidity Observations
