Abstract: Historical snow cover patterns can be derived from snow water equivalent (SWE), one of the routine outputs from reanalysis systems and offline land-surface schemes. This approach produces snow maps that are consistent with historical temperature and precipitation conditions and which complement existing satellite-based snow cover records. The NOAA snow cover extent climate data record (NOAA CDR) is the longest satellite-based snow cover record, though there is ongoing debate surrounding the apparent positive trend in its October snow cover. We show that across a range of plausible snow cover trends that we derive from snow model output, the positive October snow cover trend cannot be reproduced unless the snow detection sensitivity changes over time. This explanation is consistent with some of the methodological advances over the long NOAA CDR history, which could explain the spurious positive trend.