MODIS Data at the National Snow and Ice Data Center: Improvements for Collection 6

For more than a decade, the National Snow and Ice Data Center (NSIDC) has archived and distributed snow cover and sea ice products derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) instruments, onboard NASA's Earth Observing System (EOS) Aqua and Terra satellites. As the MODIS science team studies and refines the algorithms that generate these products, the Earth Observing System Data and Information System (EOSDIS) periodically reprocesses an entire collection of MODIS data and produces a new suite of products incorporating the latest enhancements. Collection 6 represents the next revision to NSIDC's MODIS archive, mainly affecting the snow-cover products. Based on scores of journal papers and workshop proceedings, Collection 6 specifically addresses the needs of the MODIS science community by targeting the scenarios that have historically confounded snow detection and introduced errors into the snow-cover and fractional snow-cover maps. Although MODIS snow-cover maps are typically 90 percent accurate or better under good observing conditions, Collection 6 uses revised algorithms to discriminate between snow and clouds, resolve uncertainties along the edges of snow-covered region, and detect summer snow cover in mountains. Furthermore, Collection 6 applies modified and additional snow detection screens and new Quality Assessment protocols that enhance the overall accuracy of the snow maps compared with Collection 5. Finally, Collection 6 introduces new MODIS snow products, including a daily climate modeling-grid, cloud gap-filled (CGF) snow-cover map. The CGF algorithm generates cloud-free maps by using the most recent clear observation of the surface when the current day is cloudy, and tracks cloud persistence to account for uncertainties created by the age of a snow observation. By considering prior days, the CGF dramatically increases the number of observable grid cells and can potentially improve the accuracy of other snow-cover products that require cloud-reduced or cloud-free observations. References Riggs, G. and D. K. Hall. 2012. Improved Snow Mapping Accuracy with Revised MODIS Snow Algorithm. Presented at the 69th Eastern Snow Conference, Claryville, NY.