A New Approach to Snow Climatology with the Winter Storm Severity Index

The Winter Storm Severity Index (WSSI) is achieved through the use of Geographic Information Systems (GIS) by screening the official National Weather Service (NWS) gridded forecasts from the National Digital Forecast Database (NDFD) for winter weather elements and combining those data with non-meteorological or static information datasets (e.g., climatology, land-use, population) to create a graphical depiction of anticipated overall impacts to society due to winter weather. The WSSI currently consists of six components: snow amount, snow load, blowing snow, ground blizzard, snow load and ice accumulation. The WSSI has been developed to have a two-fold focus. One is as a tool to assist NWS operational forecasters in maintaining situational awareness of the possible significance of weather related impacts based upon the current official forecast. The second is to enhance communication to external partners, media and the general public of the expected severity (potential societal impacts) and spatial distribution of winter weather.

The WSSI has been running at Weather Prediction Center (WPC) since October, 2017. Upon investigation of the calculations of severity levels, the snow amount climatology that helps determine snow amount severity was seen as an area that could be improved. This was due to the limited amount of reference stations used in the initial climatology calculations. A statistical approach was employed that utilized a newly developed snow climatology based upon one and two-day maximum snow amounts. Average and standard deviations of the one and two-day maximum snowfalls were calculated for all 6,055 stations that reported snow within the Historical Climate Network (HCN) for all available data between 1965 and 2015. The HCN stations are a subset of the NOAA Cooperative Observer Program (COOP) Network with sites selected according to their spatial coverage, record length, data completeness, and historical stability. A tool developed by the Environmental Systems Research Institute (ESRI) called EBK Regression Prediction was used to interpolate the data between stations. The tool uses Empirical Bayesian Kriging and Ordinary Least Squares (OLS) regression on contributing independent variables to create a more accurate interpolated value than than a regression model or kriging interpolation could provide on their own. The interpolated grids were used to create new critical thresholds for the snow amount element within the WSSI. The methods used to create the new snowfall and a comparison of the old and new thresholds will be presented.