Cloud context-based onboard data compression

A simplified cloud mask algorithm suitable for present-day onboard processing is developed based on the operational MODIS cloud mask algorithm. The code is reduced to less than 45% of its original size while retaining 90% of the accuracy. Clear sky data compression is tested using our simplified cloud mask followed by application of the CCSDS lossless compression algorithm. For most of the channels, compressing the cloud masked MODIS L1B data can reduce the data volume by about 40% compared with compression without using cloud mask. The cloud mask was also applied to L1A data with comparable or even better data reduction performance depending on local data filling schemes. The cloud mask itself can be compressed to less than 0.5 bit-per-pixel. This paper summarizes our approach and provides results in detail.