Temporal filtering methods for operational snow covered area mapping

Single-day snow covered area (SCA) products are incomplete and often inadequate representations of ground conditions due to short term variation in cloud cover, snow cover, and sensor geometry. We discuss the applicability, benefits, and drawbacks of two temporal filtering methodologies designed for use with moderate resolution SCA data: the cloud-gap-filled (CGF) product (Hall et al., 2010) and the time-domain-filtered (TDF) product (Morriss et al., submitted, 2015) using both MODSCAG and NDSI-based SCA products (MOD10A1, VIIRS Snow Cover EDR). These filtering algorithms both use previous days' data to estimate the current SCA value of each pixel, though the TDF algorithm requires temporally persistent snow cover to map snow in the output. We investigate how the filtered products relate to the daily inputs as well as to each other and assess their accuracy based on manually derived and quality-controlled snow maps produced by the U.S. Army Corps of Engineers Cold Regions Research and Engineering Laboratory (USACE CRREL). We show both algorithms successfully remove the majority of cloud cover from the daily scenes, demonstrate how TDF successfully limits false positives caused by snow/cloud confusion in MODSCAG, and discuss the further development and implementation of these methodologies to produce a cloud-sparse, global, daily SCA product using VIIRS.