Estimating daily average net radiation in Northern Mongolia

Abstract

ABSTRACT Net radiation is a key component of surface radiation balance and has a strong influence on hydrological processes via evapotranspiration. In this study, daily average net radiation (Rnmean) for all-sky conditions was investigated as a function of the estimated daily average global radiation (GRmean) during the summers of 2011 and 2012 (128 days total) in the Sugnugur Valley of Northern Mongolia. We present a simple alternative remote sensing approach that considers factors such as topography, cloud fraction, cloud optical thickness and surface albedo. First, a geometric model for the simulation of daily average global radiation (GRCS:mean) for clear-sky conditions was applied on a daily basis. It considers topographical effects, such as slope, azimuth and elevation. GRmean was then derived for all-sky conditions by coupling the averaged atmospheric products of MODIS. Finally, Rnmean was obtained as a function of the simulated GRmean using the linear regression parameters found at a permanent observation site. The results were validated with the data from a nearby temporary observation site. The root mean square errors (RMSE) were 44 and 52 Wm−2 for GRmean and 18 and 25 Wm−2 for Rnmean at the two different sites. This methodology requires few observations and offers a simple means for estimating GRmean with high spatial (30 m) and temporal (daily) resolution under any sky conditions in the absence of ground measurements. Furthermore, Rnmean can be modeled from the simulated GRmean at regional or watershed scales where ground observations exist at one site at least.