Bright sunshine duration in relation to precipitation air temperature and geographic location

Bright sunshine duration (BSD) is of singular importance for estimating solar irradiance, and its data base is small in comparison. This paper reports a generic algorithm that captures global variation in monthly BSD data in relation to temperature, precipitation and geographic location. The algorithm depicts BSD in terms of reduction from daylength as a negative exponential function of standard station pressure, with the exponential reduction rate as proportional to the product of a series of multiplicative functions reflective of global generalities and regional idiosyncrasies. The algorithm is parameterized byregression fitting to monthly climate normal data for 729 stations worldwide. It accounts statistically for 85% of the variance in the BSD data, with a root mean square error of 1.0 hr, or 15% of the data mean. The data fitting tends to be least robust for tropical humid climates or for tropical and subtropical monsoonal Asia. The spatially-based algorithm projects with a reasonable accuracy 5-yr sequential monthly BSD data for five stations representative of the climate regimes in the conterminous United States: the absolute error is within +/-1.5 hr for 70% to 93% of the 60 monthly mean BSD values for each of the stations. [ABSTRACT FROM AUTHOR]