During Fall 2004, Version 3 of the functional prototype Maintenance Decision Support System (MDSS), intended to serve as a primarily wintertime decision support tool to state and local highway maintenance departments, was released by a consortium of institutions (NCAR’s Research Applications Program, NOAA’s Forecast Systems Laboratory, MIT’s Lincoln Laboratory and CRREL) for operation during a second winter field trial phase. This version of the MDSS includes a heuristic “fuzzy logic” algorithm for diagnosing and predicting blowing snow that utilizes observations and numerical forecasts of hydrometeorological variables, and was developed from climatological considerations related to blowing snow events within the state of Iowa.

In this paper we present preliminary results of an ongoing effort to validate this algorithm for a broader geographic area as befits the intended MDSS application as a prototype national decision support tool. Given the relative lack of blowing snow data and the often localized nature of blowing snow events, such validation is a challenging enterprise, particularly in terms of application to the road weather environment. As such, we are performing the validation in two stages and with a geographic focus designed to provide the most robust validation statistics possible.

In particular, phase I of the validation involves comparison with standard manned and automated hourly NWS observations in the northern plains, Midwest, and Alaska. Current plans call for this validation to focus on winters during the 1994-2004 period. Data from supplementary sources, including SNOTEL and satellite observations, and surrogate variables such as visibility, are utilized to the extent necessary to estimate occurrences of blowing snow where it is not explicitly reported. Current plans call for Phase II to focus more directly on the road weather environment by considering, individually and as a collective whole, Road Weather Information System (RWIS) sites and bring to bear more specific geospatial information and surrogate data where required.

At the conference we will present our methodology for this work along with early results from Phase I of the validation. In addition, we plan to present any available results corresponding to Phase II of the validation effort, as defined above, as well as preliminary recommendations based on our findings.