Parameterization of solar fluxes over mountain surfaces for application to climate models

Abstract

[1] On the basis of 3‐D Monte Carlo photon tracing simulations, we have developed a parameterization of solar fluxes over mountain surfaces by means of the multiple linear regression analysis associated with topographic information, including elevation, solar incident angle, sky view factor, and terrain configuration factor. For clear skies without aerosols and clouds, the regression equation for the direct flux can explain more than 98% of the variation in which the solar incident angle is the dominant factor, except when the Sun is very low or at zenith. About 60% of thevariation in the diffuse flux is predicted bythe regression equation in which the mean elevation, sky view factor, and solar incident angle are key factors. The terrain‐reflected fluxes, proportional to the surface albedo, are well correlated with the terrain configuration factor with more than 80% of the variation that can beexplained. Thecoupledfluxes involveintricate interactions, andtheregressionanalysis is less satisfactory in cases of low albedo values. However, over high‐albedo surfaces, the terrain configuration factorbecomes most dominant, leadingtoasignificant improvement in regression performance. In these analyses, a surface albedo invariant with wavelength has been used. Using a region over the Sierra Nevada as a testbed, the preceding regression parameterizationshavebeenspecificallydevelopedsothatthefluxesevaluatedfromthe3‐D Monte Carlo model over intense topography can be used as a perturbation term to correct those computed from the plane‐parallel counterpart, commonly used in regional climate models and GCMs.