Research Objectives

The specific work of the Center is the application of fine-scale analysis and prediction techniques associated with pavement frost, pavement precipitation accumulation, blowing/drifting snow, and roadway visibility as part of an improved roadway weather management system, traveler information system and operations and maintenance activities. Many of the required techniques necessary to complete this task have been researched and developed through other non-surface transportation efforts. The use of modern meteorological data assimilation coupled with data quality control and correction techniques can synthesize the existing RWIS data with other surface-based meteorological observations to provide an enhanced representation of weather conditions. The proposed work will perform this data assimilation activity to provide a more robust and complete roadway weather management system. The proposed work will also incorporate into this data assimilation effort remotely sensed meteorological data that will provide greater insight into the presence of precipitation, its rate of accumulation and its redistribution due to the wind.

Further, using mesoscale modeling in combination with statistical and heuristic prediction methods the prediction of future pavement conditions resulting from spatial variations in weather conditions will be accomplished. Development of a roadway specific modeling system that incorporates the asymmetric dimensions of the roadway system will be evaluated as a potential method for reducing further the 10-kilometer resolution to be used in the initial project pavement condition work. The purpose of this roadway specific model is to overcome the problems with present forecasting methods in the roadway system that are based upon the conformance of coarser spatial model or forecast system information onto the roadway environment. Having a prediction system designed to conform to the geometry of the roadway permits more explicit prediction systems to evolve. The use of adaptive grids and finite element methods are a logical approach to this roadway modeling environment. From this modeling environment more detailed simulations/predictions could be driven by the boundary conditions derived from the existing model environments.

In addition, a reliable multi-dimensional snowdrift prediction model will be developed that provides support for road coverage (pillow drifts, finger drifts, etc.) and support for driver visibility prediction. Present models by the U.S. Army Corps of Engineers in use today (SNTHERM) are non-distributed models (only computed for a vertical column above a point). Using the roadway modeling system suggested above, this snowdrift model should be capable of being driven by improved mesoscale and misoscale prediction models. The modeling of spatial distribution of blowing and drifting snow throughout a road corridor from this model would also provide valuable information for predicting driver visibility (considering both the driver height levels of automobiles and trucks). Especially important will be the need to understand the roadside land-surface characteristics and the cross-section geometry of the roadway. This information will be valuable in a 0-6 hour prediction for the purpose of providing travel information and maintenance support.

Finally, an improved road condition model will be constructed that takes into consideration not only the heat balance condition, but also mass balance that includes the impacts of road surface contaminants. An existing operational mass balance model will be utilized as the basis in the final pavement condition prediction efforts. Again, as before, much of the detail lies in knowing what actually is found on the road surface. Another important element in this mass balance road condition model is the impact of traffic on the distribution of snow and ice on the road surface. Pavement level modeling of traffic impacts on the observed road surface will be an important element in this completed model, but are beyond the present scope of work to be performed.

This pavement conditions system, known as the Advanced Pavement Condition Analysis and Prediction System (APCAPS), will provide solutions to the spatial variability of pavement problems associated with too few RWIS observations and the inability of RWIS to predict future conditions. This APCAPS information will be integrated with other ITS systems requiring improved pavement condition data and information to promote improved safety, mobility and cost effectiveness. By leveraging the proposed efforts with other existing and proposed ITS efforts, it is expected that not only an economy of scale will be achieved, but that the time to deploy will be dramatically reduced.

The APCAPS will be part of a collection of ITS initiatives either already in place or initiatives that are soon to begin. A portion of the work will build upon previous ITS deployment activities at the University of North Dakota through activities with the North Dakota and South Dakota Departments of Transportation. The States of North Dakota and South Dakota were early developers and adopters of advanced traveler information systems in the late 1990s. Their efforts were part of a FHWA funded ITS research and deployment effort at the University of North Dakota known as the Advanced Transportation Weather Information System (ATWIS). The success of ATWIS from 1996 through 2001 resulted in the ATWIS methodologies leading to the first statewide 511 system deployment in Nebraska in late 2001. Since then both North Dakota and South Dakota have transitioned their early ATWIS systems, referred to as #SAFE, to a 511 configuration. The basis of this system is comprised of a short-range, site-specific weather forecast and the latest reported road conditions as determined by maintenance personnel and/or state highway patrol personnel. And although the weather forecasting efforts include the use of Road Weather Information System (RWIS) observations of pavement conditions, the RWIS pavement condition information is not provided to the traveler in the 511 systems due to its lack of representativeness beyond its observation location and its relevance only as an observation and not a prediction of future conditions. The proposed APCAPS will be integrated with the existing 511 systems in both states to enhance the road condition content.

Besides supporting 511, APCAPS will improve the utilization of existing and future Road Weather Management Systems (RWMS). In order to provide a reliable pavement condition observation system using existing methods requires a massive fleet of trained observers continually traveling the highway and/or freeway system or an increase in RWIS sensor locations of many orders of magnitude beyond the number presently deployed. The economic feasibility of either of these solutions is not realistic nor would such a deployment provide information on the future pavement conditions that would be encountered by travelers or require removal by maintenance personnel. The proposed activity, as will be described in more detail in a later section, will draw upon the existing RWMS, in the form of RWIS observations across North Dakota and South Dakota, to build a more extensive RWMS that can be utilized more efficiently.

The APCAPS will provide much of its information to 511 via a sophisticated road condition reporting system (RCRS) presently being designed for the South Dakota Department of Transportation by Meridian Environmental Technology, Inc. This RCRS, which will also be deployed in North Dakota during 2004, will provide the necessary support for data and information delivery following accepted ITS Standards. As a result, the information from APCAPS will be able to flow to all applications supported by the RCRS. And because APCAPS will follow ITS data standards, any RCRS following these same ITS standards will be capable of utilizing the information of APCAPS. This will permit a desired broader adoption and diffusion of the resulting ITS technologies of APCAPS beyond North Dakota and South Dakota.

The final existing ITS system in which APCAPS will be included is MDSS. As described in Section 1, the lack of quality pavement condition information having fine spatial resolution has been a problem within current MDSS development. To provide the required pavement condition information, either analyses or predictions, has been beyond the scope of work for the present MDSS initiatives. The proposed project will provide a valuable gap closer in the required information needed to make MDSS successful. The research and development to be conducted as part of the deployment process for APCAPS will prove to be beneficial for future MDSS deployments.