A self-adaptive approach for producing clear-sky composites from VIIRS surface reflectance datasets

With the launch of the Joint Polar Satellite System (JPSS)/Soumi National Polar-orbiting Partnership (S-NPP) satellite in October 2011, the need for the operational monitoring of terrestrial processes at the regional and global scales led to the expansion of terrestrial remote sensing products (e.g., the clear-sky composited surface reflectance products) generated from the Moderate Resolution Imaging Spectroradiometer (MODIS) into the JPSS/S-NPP mission using the new Visible Infrared Imaging Radiometer Suite (VIIRS) data. Seamless cloud composites are usually generated using a single criterion without an explicit consideration of phenological variations among different surface types. However, because the spectral signals of many surface types change dramatically due to seasonal variations, the single-criterion compositing methods are only effective for specific surface cover conditions. This study proposed a new self-adaptive compositing approach (SA-Comp) to produce global terrestrial clear-sky VIIRS surface reflectance composites. The proposed approach employs contextual spectral and temporal information to determine the surface cover conditions within a pre-defined temporal window, and adaptively selects the most suitable criterion. A comprehensive evaluation of the SA-Comp approach was conducted by comparing it with the maximum NDVI (MaxNDVI), minimum Red (MinRed) and maximum ratio (MaxRatio) compositing schemes, and with the MODIS and VIIRS composited surface reflectance products. The results, including visual representations and temporal profiles, revealed that the SA-Comp approach outperformed all of the other methods. The results also highlighted that the SA-Comp approach is more feasible and effective at compositing global VIIRS data and has great potential for regional, national and even global terrestrial monitoring.