Each year, adverse weather along roadways results in over 7,000 fatalities, 673,000 injuries, nearly $9.5 billion in congestion delays, and costs shipping companies $3.5 billion. Currently, decision support systems recommend road surface treatments for snow removal operations by incorporating out... Show moreEach year, adverse weather along roadways results in over 7,000 fatalities, 673,000 injuries, nearly $9.5 billion in congestion delays, and costs shipping companies $3.5 billion. Currently, decision support systems recommend road surface treatments for snow removal operations by incorporating output from pavement temperature models. However, current pavement temperature models issue point forecasts that fail to describe the conditions between the points. A forecast from a gridded pavement model is a possible solution to this problem. To produce this forecast, the sensitivity of the pavement temperature when stratified by different materials (i.e., concrete and asphalt) and their thicknesses must be assessed. This analysis used eight Road Weather Information System (RWIS) stations in the greater Denver, Colorado metropolitan area and four RWIS sites at Denver International Airport. Uniform radiation was assumed using a cross-referencing method between five-minute daily weather observations and satellite-surface composite maps. Sixteen total case studies were selected for day and night with diverse weather conditions. Concrete roads generally averaged 3.2° C warmer than asphalt roads and therefore will have to be modeled separately. Elevated (e.g., overpass) and non-elevated surfaces were also compared, and elevated roads tended to be 3.3° C cooler than non-elevated; therefore, a condition protocol for the elevated site will subtract 3° C from the standard forecast. This work lays the foundation for additional analyses with a more robust data set to confirm initial results observed from these data, the results of which may be used to develop and tune gridded pavement model forecasts. Show less
