Lawrence Berkeley National Laboratory

Fundamental process understanding and description of heat, mass, and momentum exchanges across the land-atmosphere interface in model boundary forcing parameterizations is critical to the simulation of near-surface soil moisture dynamics (e.g., bare-soil evaporation). This study explores the sensitivity of a continuum-scale porous media heat and mass transfer model to the spatial-discretization length-scales (i.e., spatial-resolution) of near-surface atmospheric data; the goal is to determine how much data are needed to force the model and adequately capture evaporative water losses and subsurface state variable distributions. The requisite atmospheric forcing data were taken from the high-resolution, precision bare-soil evaporation experiments of Trautz et al. (2018, https://doi.org/10.1029/2018WR023102). Simulation results demonstrated that shallow subsurface mass and heat transfer dynamics can be adequately captured with forcing data averaged over large length-scales, or a minimal number of measurements, provided that soil conditions are properly described. The soil moisture spatial distributions were found to be insensitive to horizontal variations in the forcing data. The model failed to capture small-scale trends observed experimentally; this did not impact the accuracy of total evaporative water loss estimates however. These results indicate that in future physical experimental efforts conducted at 1–10-m length-scales, there is no need to focus on the generation of high-spatial resolution atmospheric measurements—time and effort would be better spent in characterizing soil conditions and properties. Even though a theoretical foundation was not provided to directly extrapolate this work to the field scale, these findings have practical value in designing field data collection strategies.