Estimating global solar radiation from common meteorological observations in western Canada

Abstract

The interest in using crop growth simulation models for estimating large area yields in western Canada has led to a requirement for daily values of solar radiation on an historical and a real time basis. Because such data are usually not readily available, an equation was developed which relates solar transmissivity (ζ) (the ratio of incoming global solar radiation at the earth surface (Q) to solar radiation at the top of the atmosphere (Q0)) to daily observations of maximum and minimum air temperature and total precipitation (P):[Formula: see text]where a, b, c and d are empirical coefficients which vary with time of year and ΔT is the range in daily temperature extremes. During the late fall-winter period, correlation coefficients between observed and calculated transmissivities were less than 0.5 with relative large root mean square errors (RMSE). However, during the growing season when the equation would be of most use, correlation coefficients were 0.7 or higher with RMSEs of 0.12 or lower. The coefficients a, b, c and d were found not to be site-specific during the growing season. No significant differences were found between wheat yields estimated with observed solar radiation and those estimated with the calculated solar radiation from the equation. Key words: Solar radiation, transmissivity, crop growth modelling, wheat yield