The weather portion of the system is being provided by Colorado Research Associates. A complete data collection, data analysis and numerical weather prediction system has been installed for use within Foretell. Standard meteorological data sets are collected via the National Weather Service Noaaport satellite system on-site in Boulder. Additional local data is provided from IDOT using automated RWIS mesonet sites located along many Iowa highways. The data is assimilated hourly, using the Local Analysis and Prediction System (LAPS) originally developed at the NOAA Forecast Systems Laboratory, onto a 10-km horizontal grid with 21 isobaric levels. Analyses are then available as initialization to the Penn State/NCAR Mesoscale Model (MM5) numerical model. The model grid uses a double nest configuration with a 30-km grid spacing on the outer grid, and a 10-km inner grid centered on Iowa with coverage over much of the Midwest. Model forecasts are generated four times per day out to 30 hours. Web-based images are created from standard meteorological analysis and forecast fields.
An important user of the system is the IDOT highway department who has snow removal responsibilities. These responsibilities include chemical treatment decisions to the pavement surface. MM5 forecasts are interpolated to each mile of state, US, and interstate highway in the Midwest. This data, in addition to being available for display, is input into a CRC-developed "predictive road model" that has knowledge of ground and pavement characteristics. The road model provides predictions of pavement temperature and surface snow/ice build-up which are then available as a support tool for highway chemical treatment decisions.
The presentation will include a discussion of the data collection, data analysis, numerical weather prediction, and road model techniques. Examples of the Foretell system will be illustrated with a winter-time case study analysis.