Abstract
The simple physical model to estimate surface insolation from GOES data (described in Gautier et al., 1980) has been improved through some modifications to existing physics (Rayleigh scattering and water vapor absorption), and also the inclusion of ozone absorption, previously neglected. An empirical correction for clouds smaller than the GOES sensor field-of-view has also been introduced. The resulting model is more physically realistic than the old one and requires no additional computer time. Tests indicate that with minimal tuning, the error in daily insulation estimates has been reduced by ∼1% compared to the old model. Removal of systematic error resulted in an additional 0.2% improvement.
