Setting of cloud albedo in the atmospheric correction procedure to generate the ocean colour data products from OCM-2

Abstract

Cloud masking is one of the primary and important steps in the atmospheric correction procedure in particular to coastal ocean waters. Cloud masking for ocean colour data processing is based on the assumption that the water reflectance is close to zero in the nearinfrared (NIR). This is valid over the open ocean, but coastal waters may have a higher NIR reflectance due to the presence of suspended sediment concentration and non-maritime aerosols. As the result, more scattering is observed in the water leaving radiance and hence sometimes the cloud-free pixels are also classified as clouds, leading to data loss. In this present study, we have analysed cloud albedo value for masking the clouds using the default cloud masking threshold (0.027) set for processing the global ocean colour sensors like SeaWiFS and MODIS and now OCM-2 in the SeaWiFS Data Analysis System (SeaDAS). The default cloud masking value of SeaDAS over masks the clouds and sometimes the turbid coastal water pixels are also getting masked when implemented over the COM-2 imageries. Hence, an exercise has been carried out to set the cloud albedo value for OCM-2 sensor and the albedo values were tested with the lower and higher values ranging from 0.015 to 5.0 as the threshold and from the analysis it is found that 0.07 is the optimum value for Ocean Colour Monitor-2 (OCM-2) sensor for masking the clouds. Using this threshold, maximum valid ocean pixels are achieved compared to the default cloud albedo mask with the value 0.027. This has been checked with both the Local Area Coverage (LAC) and Global Area Coverage (GAC) products of OCM-2 and the results were also compared with the MODIS data. Hence, it is recommended that 0.07 is the optimum cloud masking threshold for OCM-2 sensor