Comparative study of statistical and artificial neural network's methodologies for deriving global solar radiation from NOAA satellite images

Abstract

AbstractThe use of satellite data to estimate global solar radiation (G) at ground level has become an effective way for a large area with high spatial and temporal resolution. The statistical approach is a widely applied procedure for this task. The first objective of this study was to examine the potential of this approach for deriving instantaneous G from NOAA–AVHRR satellite data for the atmosphere of semi‐arid environment of Iran. The second objective was to apply artificial neural network (ANN) to the estimation of G from advanced very high resolution radiometer (AVHRR) images. A Comparison between these two methods was the last objective of this study. A total of 661 images of NOAA–AVHRR level 1b, covering the area of this study were collected from the Satellite Active Archive of NOAA. The results demonstrated that the use of ANN model gave better estimates than the statistical technique. Root mean square error and R2 for the comparison between observed and estimated G for the tested data using the proposed ANN model are 56.95 W m−2 and 0.90, respectively. For the statistical approach method these values are 68.33 W m−2 and 0.86. Copyright © 2012 Royal Meteorological Society