An Improved Approach for Automated Cloud Detection in VIIRS Imagery

The presence of clouds in the atmosphere can affect many weather and climate applications from sea surface temperatures to boundary layer behavior. Knowing whether or not there is cloud cover and how it changes with time over a region can help the operational weather community with future forecasts, as well as providing more information about the current state of the atmosphere and its processes. A version of the Bispectral Composite Threshold (BCT) approach used in past applications with MODIS imagery has been adapted for VIIRS and subjectively tuned to optimize performance in various seasons over the continental United States. The BCT approach has been simplified to include only three cloud tests using the shortwave and longwave infrared channels on VIIRS. A ground “truth” data set generated from manual determination of cloud cover extent based on multiple remote sensing scientists has been used to statistically evaluate how well the VIIRS cloud mask algorithm performs under a variety of conditions. A comparison has also been made to the VIIRS cloud mask produced by the NESDIS Interface Data Processing Segment (IDPS) and used in various weather applications. The results of this comparison will be presented at the conference.