Enhanced snow and ice identification with the VIIRS cloud mask algorithm

Abstract

New procedures have been developed to help identify snow and sea ice with the Suomi-National Polar-orbiting Partnership (S-NPP) Visible Infrared Imager Radiometer Suite (VIIRS) Cloud Mask (VCM) algorithm. The accurate detection of snow and sea ice is necessary in order to apply the correct spectral tests needed to detect clouds and make accurate cloud confidence classifications. During the VCM Calibration Validation activity, it was found that the procedures in place at the time of the satellite launch occasionally produced four types of misclassifications: (1) snow and/or ice surfaces in dry atmospheric regions misclassified as clouds, (2) multi-layered clouds in humid regions misclassified as snow, (3) low-level clouds with glaciated tops misclassified as sea ice, and (4) frozen lakes not classified as ice. The new procedures presented in this article use data collected in the VIIRS mid-wavelength region, i.e. both the 3.7 μm and 4.0 μm bands, as well as the 12.0 μm IR band to eliminate all four types of misclassifications. The results demonstrate that split window, mid-wavelength IR imagery provides valuable information for developers of automated cloud classification algorithms as well as those who generate sea ice analyses in support of ocean navigation during polar wintertime conditions. It is concluded that collecting data in these mid-wavelength IR bands should be considered part of any future satellite sensor designed for environmental monitoring.

Acknowledgements

The results presented herein were created by Northrop Grumman under AMS Contract 1707, NPP CalVal Mission Support.