NOAA Interactive Snow Charts (IMS), 1997-2024: Is the snow mapping lag decreasing over time ? Implications for operational applications and snow climatology.

NOAA interactive snow and ice cover charts of the Northern Hemisphere present a key input to operational weather prediction models and are widely used  in climate change studies. Since 1997 charts are generated daily within Interactive Multisensor Snow and Ice Mapping System (IMS). When creating a snow map, analysts use the previous day map as a first guess and update it manually using satellite imagery in the optical bands. Because of clouds obscuring the view and limited time allocated for the analysis only a portion of the entire land area may be closely examined on a given day. Otherwise the snow cover is assumed unchanged since the day before. Inability to update the map over its entire domain on a daily basis results in a delayed reproduction of snow cover dynamics in the map and, hence, in degraded map accuracy. Over time various image and data analysis tools as well as several auxiliary snow cover datasets have been added to the IMS to facilitate the work of human analysts. The spatial resolution of the maps has substantially increased, and it is believed that the snow mapping time lag or the map timeliness might have also improved owing to the system enhancements. However, the extent of this latter improvement and whether it actually occurred remains uncertain. 
In this work we sought to estimate the time interval between consecutive updates of the IMS snow cover map (or the revisit time interval) and to understand whether the frequency of updates and, hence the map timeliness, has improved over time.  IMS snow maps at 24km spatial resolution over the 1997-2024 time period have been used. We examined daily snow extent records over several relatively small test regions and estimated the time interval between consecutive updates of the map. The focus was on the period of the most active seasonal snow melt when the rate of the snow cover retreat required daily updates of the snow map. We have found that at the inception of the IMS system in 1997, the mean frequency of updates of the IMS snow map fluctuated between once every 4 - 6 days rather than daily. By 2024 the revisit time interval dropped to about 2 days with most of the decrease occurring in the first decade of the century. With the frequency of updates improved by 2 to 4 days, the mean time lag to reproduce snow cover dynamics in the IMS snow cover map, or the map latency decreased by 1-2 days.  Somewhat greater improvement in the update time interval and, hence, in the snow map latency was observed over mountainous areas as compared to predominantly flat terrains. 
While decreasing snow mapping time lag and, hence, improved timeliness of IMS snow maps is certainly beneficial for operational applications, it poses challenges for the climatological use of the dataset. In the presentation we estimate and discuss the effect of improved timeliness of the product on the snow extent and snow phenology trends inferred from the IMS snow dataset.