In order to facilitate the proper snow and ice control operations, it is essential to identify and forecast road surface temperature and/or road condition. The Civil Engineering Research Institute for Cold Region has directed efforts towards developing a road surface temperature prediction model, utilizing a heat balance method which can account for the effects of vehicular traffic. Results obtained from the development and testing of this model were presented at the 85th Transportation Research Board Annual Meeting. These results demonstrated that the model accurately predicted trends in surface temperature variation, but consistently underestimated the actual temperature value. The error between the predicted and actual temperature values was especially large during the nighttime. It was believed that failure to account for outgoing longwave radiation from structures along the roadway could be the source of this discrepancy. Therefore, our research group attempted to improve the model to account for the shielding effect of roadside structures, as well as for outgoing longwave radiation from them, in its calculation of net atmospheric radiation. Besides, a long and short wave radiation sensor was employed to measure actual levels of downward longwave radiation. By doing so, the root mean square error in predicting longwave radiation levels was shown to be improved from 39.5 W/m2 to 17.8 W/m2. The surface temperature prediction error was then improved substantially by an all-day average of 1.5 degree Celsius. Average error for the daytime and nighttime were improved by 1.4 degree Celsius and 1.7 degree Celsius respectively.
Samenvatting