Snow drifting: Drifting snow creates dangerous driving conditions,
particularly when the drifts are intermittent. The users continue to seek
solutions that provide guidance on drifting snow. The feasibility of using
snow drift models in real time should be analyzed.
The MDSS prototype improved significantly between the first and second field
demonstration periods. The end users, primarily Iowa DOT maintenance supervisors,
indicated that they felt much more comfortable with the system and that the treatment
recommendations were much closer to the actual treatments. The 2003-2004 winter field
demonstration included a broad range of weather and road conditions and provided an
excellent dataset for analysis.
Weather and road condition prediction on road scales is complex and poses significant
challenges for decision support systems for winter maintenance operations. The MDSS
project has been very useful in highlighting areas that need attention as well as
capabilities (techniques, methods, features, and functions) that are ready or nearly ready
for transition to operational use.
There were several successes this year. They include:
a) An improved understanding of numerical weather prediction, particularly
mesoscale modeling, and its current capabilities and limits for decision support
applications such as the MDSS.
b) The success of the Road Weather Forecast System's data fusion methods (e.g.,
model output statistics, dynamic weighting, forward error correction, etc.) in
providing additional weather forecast skill above and beyond the individual
c) The use of direct insolation (short wave solar radiation) data from the mesoscale
models instead of calculating the insolation from cloud field data improved the
road temperature predictions.
d) The use of "hot start" models improved precipitation forecasts. Maintenance
personnel have indicated on several occasions that a better prediction of
precipitation start and end times would be a major benefit.
e) The revised Rules of Practice code that included logic to characterize the pre-
storm, during storm and post storm conditions and covered more weather and
road condition scenarios, was a significant success.
f) The provision of probabilistic information on precipitation type allowed users to