Estimating south slope irradiance in the Arctic—A comparison of experimental and modeled values

Abstract

The limited availability of experimental data on inclined surfaces often requires that slope irradiance be modeled using data obtained for a horizontal surface. In this article, estimates of irradiance on a south facing 80° inclined surface obtained using the models of Hay[1] and Klucher[2] are compared with experimental results obtained at Alert, N.W.T. (82.50°N). Subdivision of global radiation into direct and diffuse components was achieved using the methodology of Liu and Jordan[3]. This is improved by using empirical constants derived for high-latitude areas. Surface reflection has been treated assuming both isotropic and anisotropic reflection. Results indicate the models provide reasonably accurate estimates of slope irradiance from April to August when the solar elevation angles generally exceed 6 to 8°. During March and September, periods of low maximum solar elevation angles, the models overestimate slope irradiance. The failure of the models to adequately predict irradiance at low sun angles on the 80° incline appears related to a break-down of the relationship between sky condition and atmospheric transmissivity, ( K t ). At low solar elevation angles, the portion of global radiation that is attributed to being direct is overestimated. This results in overestimates of slope irradiance under these conditions.