Evapo-sublimation Variability Across a Forest/Meadow Transition Zone

Abstract:

Evapo-sublimation processes in snow-covered forested regions may constitute up to 25% loss of seasonal snowfall and significantly affect the winter water budget. However, this hydrologic component is difficult to measure and model, leading to a lack of understanding of its importance. Advances in measuring and modeling snow in complex, forested terrain have led to key improvements in our knowledge of snow depth and simulations of snowpack evolution. However, basin scale models are not able to resolve energy and moisture fluxes at forest/open interfaces and at canopy level where large gradients in state variables drive sensible heat (SH) and latent heat (LH) fluxes. Here, we describe key findings from a field experiment focused on measuring evapo-sublimation across a forest density gradient in the California Sierra Nevada. The Sagehen Winter Experiment spanned two winter seasons (2019 and 2020). In 2019, we deployed five eddy covariance (EC) flux towers across a transect between dense forest and open meadow that was oriented along the prevailing wind direction to measure localized contributions to SH and LH fluxes. In 2020, instrumentation was consolidated onto three towers to quantify vertical flux divergence terms. Results indicate large differences in measured fluxes across horizontal distances on the order of tens of meters and vertical distances on the order of meters. Clear-sky evapo-sublimation rates increased with wind speed and were often maximized at midday, causing a local minimum in mixing ratio. Relative to the open area site, the presence of trees increased ground-level snow deposition at forest margins but decreased direct deposition of snow in dense forest. Evapo-sublimation rate was greatest in snow-covered canopy but was constrained by diminished winds, melt and throughfall. Solar irradiance on the forest canopy provided an elevated source of heat that influenced SH fluxes over the canopy and extended to open areas. SH and LH fluxes computed using the aerodynamic method agreed qualitatively with EC-derived fluxes but underestimated fluxes computed by EC methods.