Generating Global LAnd Surface Satellite incident shortwave radiation and photosynthetically active radiation products from multiple satellite data

Abstract

Incident shortwave radiation (ISR) at the surface is an essential parameter in the land surface radiation budget and in many land surface process models. Incident photosynthetically active radiation (PAR) is required by nearly all terrestrial ecosystem models. Although several global radiation products from numeric models and satellite observations have been released, their coarse spatial resolutions and low accuracy, especially at high altitude regions, limit their applications in land community. In this study, the Global LAnd Surface Satellite (GLASS) ISR and PAR products are developed based on an improved look-up table method from 2008 to 2010 at a 5km spatial resolution and a 3hour temporal resolution, the first global radiation products at such high resolutions, from multiple polar-orbiting and geostationary satellite data, including the Moderate Resolution Imaging Spectroradiometer (MODIS), the Meteosat Second Generation (MSG2) SEVIRI, the Multi-functional Transport Satellite (MTSAT)-1R, and the Geostationary Operational Environmental Satellite (GOES) Imager. The look-up table algorithm primarily considers the influences of surface elevation, atmospheric water vapor because sensitivity experiments indicate that ISR and PAR are more sensitive to surface elevation, and less sensitive to atmospheric profiles and ozone amount. The water vapor absorption has strong impact on ISR but weak on PAR. Moreover, the surface bi-directional reflectance distribution function (BRDF) is taken into account to retrieve spatially and temporally continuous surface reflectances from the geostationary satellite observations, which is an input parameter in the revised look-up table method. Ground-based measurement data from 34 sites are used to validate the improved algorithm and the GLASS products. The validation results of the instantaneous ISR and PAR products at all validation sites are notably good with coefficient of determination values of 0.83 and 0.84, respectively, and root mean square error values of 115.0Wm-2 and 49.0Wm-2, respectively. We also aggregate the GLASS ISR product for comparing with the corresponding ISCCP and CERES data at 7 SURFRAD sites and demonstrate that the GLASS ISR product is more accurate. The GLASS ISR and PAR products have been made publicly available. © 2013 Elsevier Inc.