Analyses of the new ESA CCI+ Snow cover fraction climate data record from Terra MODIS and Sentinel-3 SLSTR data

Seasonal snow is one of the terrestrial essential climate variables specified by the Global Climate Observing System (GCOS). With a coverage of about 45 to 50 Mio. km² of the global land area during the main winter season in the past decades, seasonal snow is the largest component of the cryosphere having a major impact on different processes of the Earth’s system.

Different medium resolution optical satellite data have been exploited in the past few years to monitor the seasonal snow extent on local to global scale. Most of these satellite-based products provide information on the snow viewable from space, i.e. in forested areas the snow viewable on top of the forest canopy, and many of these products provide only binary classification on snow, i.e. a pixel is either snow covered or snow free.

In the frame of the ESA Climate Change Initiative Extension (CCI+) Snow, a new climate data record (CDR) of daily global snow cover fraction maps with about 1 km pixel spacing was generated from Terra MODIS and Sentinel-3 SLSTR data for the period 2000 – 2020. The daily products of this CDR provide the fraction of snow covered area per pixel in percentage not only for all land areas, but differentiate in forested areas two thematic snow information, the snow cover fraction viewable from above, and the snow cover fraction on the ground. The retrieval method assures that the classified snow cover fraction on ground and the viewable snow cover fraction information are consistent for all observed land areas, allowing the usage of the data sets in different applications. Each daily product contains the unbiased root mean square error per observed pixel as uncertainty estimation. The CDR will be publicly released via the ESA Open Data Portal soon.

Based on the new CDR, the variability of the seasonal snow in the past 20 years is analysed, investigating in detail interannual, seasonal and monthly trends on global and hemispheric scales. The maximum global snow cover in the past 20 years shows overall a negative trend, although the derived interannual variations reach up to 5 Mio. km². The analysis of the seasonal snow extent indicates no significant trend of the maximum snow cover during the main winter season on the Northern Hemisphere (January – March) in the past 20 years. But during the onset and the melting seasons of the Northern Hemisphere, all the trends of the maximum snow area are negative, with the most negative signal in May.

We will present the method used for the generation of the new snow cover fraction CDR from MODIS and SLSTR data and discuss the results of the spatial and temporal analyses of the 20-years time series of global daily snow cover fraction products, including also analyses of the variations in the timing and duration of the snow season for selected regions in the context of the changing climate.