Development of a Novel Road Ice Detection and Road Closure System: Modeling, Observations, and Risk Communication

The Central Plains of the United States are particularly susceptible to roadway icing hazards given the predominance of frozen precipitation, freezing fog, and hoar frost within this region.  Typically, drivers are unaware of the presence of roadway ice due to its clear nature; therefore, providing drivers with adequate warning is essential to the mitigation of societal detriments resulting from this phenomenon.  In 2012, the Oklahoma Department of Transportation (ODOT) funded an interdisciplinary effort to develop a multi-faceted road ice prediction and warning network across the state of Oklahoma.  The developed system, which has emphasized cost-effectiveness and computational efficiency, is comprised of a road icing model (RIM), a roadside weather information system (RWIS), and a GIS database for effective data access and visualization. 

The RIM offers both diagnostic and prognostic assessment of road-icing risk independent of road-surface observations.  The model categorizes icing risk using a hierarchical system that utilizes: (a) observational output from the Oklahoma Mesonet and Automated Surface Observation Station (ASOS) network for diagnostic road icing risk assessment, and (b) model output from the National Weather Service (NWS) monitored National Digital Forecast Database (NDFD) for prognostic assessment.  The RWIS was developed using novel, cost-effective methods, and is comprised of stations built from off-the-shelf piezoelectric sensors, microprocessors, and wireless modules.  A stochastic method is used to determine road icing risk, which may be readily assimilated into the diagnostic portions of RIM.  Using a combination of the output from the RWIS and RIM, a customized GIS visualization software offers road-scale road icing information.  Supplementary information regarding traffic flow, population, and topography is available for rapid in-situ warning of road icing.  The combination of an observationally parameterized road icing model, an RWIS, and a GIS visualization package captures the wide range of road icing risk, while offering computational efficiency and economical deployment.