Atmospheric moisture budget estimates of regional evapotranspiration from RES‐91

Abstract

Diurnal changes in the local atmospheric moisture budget over the Canadian Prairies are computed using sequential radiosonde soundings from the 1991 Regional Evaporation Study (RES‐91). Previous attempts to estimate evapotranspiration with radiosonde data have used either similarity theory or a moisture budget, but have been confined to the boundary layer in either case. These studies, as well as semi‐empiric operational techniques which use surface‐based data, exclude the effects of moisture advection and energy exchanges between the boundary layer and the free atmosphere, assuming negligible effects on evapotranspiration. The moisture budget method adopted here includes horizontal advection explicitly, and treats vertical fluxes implicitly through a total tropospheric moisture budget. Comparison of the evapotranspiration estimates with those of other techniques are positive only when results are averaged over several days to weeks. While the advection estimates are a major source of error for the “daily” estimates in this particular study, it is shown that neither advection nor moisture flux through the boundary layer can be ignored in estimating daily evapotranspiration, regardless of the technique used. The results also suggest that evapotranspiration is more variable on a daily basis than other techniques have indicated. With an improved synoptic database now available for advection estimates, the moisture budget technique may provide an excellent ground‐truth method for fine‐tuning techniques for remote sensing of evapotranspiration, and could lead to improved parametrization schemes for both NWP models and GCMs.