Testing a Detailed Biophysical Parameterization for Land–Air Exchange in a High-Resolution Boundary-Layer Model

Abstract

In order to properly model the influence of land surface properties on mesoscale atmospheric phenomena, it is important to include physically realistic parameterizations of major biophysical processes involved. The primary influence of the surface on the atmosphere occurs via its control of the surface energy budget and the consequent turbulent exchange with the planetary boundary layer (PBL). The physical parameterization of the complex surface processes may not be confidently incorporated into a three dimensional model without first undergoing testing in a simpler, were controlled setting. It has been accepted practice to first validate the pararmeterization in a one-dimensional version of the intended parent model. The purposes of this paper are to present the results of such a validation and to provide deeper insight into a key aspect of the parameterization by presenting some sensitivity tests involving the leaf stomatal control of water vapor flux. The performance of the new parameterization in the parent model is compared to three different observational datasets characterized by widely different surface and vegetation conditions; the individual fluxes from the new model are found to simulate the observations well and to be a significant improvement compared to the fluxes from the original model. Last, the values of latent heat flux, obtained using two independent stomatal resistance formulations, are compared. For the three experimental datasets studied, the difference in predicted latent heat flux between the two formulations is less than 10 W m?2 at all times. Although sensitivity tests showed greater differences under certain circumstances, it is concluded that most of the biophysical controls that enter into the stomatal resistance formulation, but defy simple field measurements do not need to be specified with great accuracy in order to produce a prediction of latent heat flux that falls within the envelope of usual observational error.