LAUSR

Abstract Temperature affects the structural response of asphalt concrete (AC) pavement. To evaluate the performance, it is important to accurately predict the pavement surface temperature. This study develops three regression models to determine the surface temperature of an asphalt pavement using field measured data such as air temperature, wind speed, wind direction, relative humidity, and solar radiation. Three models namely, 24-h model, day-time model and night-time model were developed based on one-year continuous data. The weather data were measured from October 2012 to October 2013 at an instrumented pavement section on Interstate I-40 near Albuquerque, New Mexico. The developed regression models were validated using October 2014 to June 2015 weather data, which were not used in developing the regression models. Surface temperatures predicted using the developed regression models were then compared to those predicted using the pavement mechanistic-empirical (ME) software's default model. It is observed that the regression models are better predictors of pavement surface temperatures than the ME default models. A parametric study was also performed in this study to identify the effects of weather parameters on the pavement surface temperature prediction. The authors of this study expect that this study will be useful in pavement design especially in cold regions.