Development and validation of a general model for estimating global solar radiation from hourly , daily and monthly surface meteorological data

Kun Yang,Toshio Koike

2017-01-01

Abstract:Long-term recorded data at world-widely distributed meteorological stations provides an opportunity to construct historical solar radiation for studies on crop productivity, hydrology, climate change, and solar power-relevant designs The Ångström–Prescott model has been widely used to estimate global solar radiation from measured sunshine duration. However, this model has site-dependent coefficients, which have to be calibrated for different climate zones and elevations, because this model does not explicitly take into account radiative extinction processes in the atmosphere. In this study, we developed a general radiation model available with global data sets for turbidity and ozone from remote sensing. Its development followed the four principles below: (1) it keeps the simple form of Angstrom model while can explicitly and accurately deal with radiative extinction processes in the atmosphere; (2) model inputs are surface meteorological data (sunshine duration, air temperature, and relative humidity) so that the model can be applied easily; (3) under clear skies, the estimated solar radiation is equal to the value calculated by sophisticated spectral models; (4) at the top of atmosphere, the estimated solar radiation is exactly equal to the theoretical value of solar radiation. The model is calibrated with data in Japan, while it is successfully applied to China, Saudi Arab, and America, where have distinct climate zones and elevations. We expect the radiation output of this model could be used for cross-validation with satellite products.