S M Kamal Hossain

Deicing Performance of Road Salt: Modeling and Applications

This paper describes the results of a field study designed to quantify the effects of various factors on the snow-melting performance of salt. Many tests were conducted in a realistic environment over two winter seasons, covering more than 70 snow events, with temperatures ranging from −14°C to 3°C and snowfalls ranging from ∼0.2 to 21.0 cm. For each snow event, salts were applied to a set of test sections with specific application rates, and time series performance; condition data such as snow coverage or bare pavement status, friction, pavement and air temperature, sky view, and humidity were collected. An exploratory data analysis was performed to identify the key factors influencing the snow-melting performance of salt, such as application rate, temperature, and snow amount. A multiple linear regression model was calibrated for the relationship between the key snow-melting performance indicators: bare pavement regain time and various influencing factors. The calibrated model was then applied to determine the minimum amount of salt required for achieving a given level of service under specific weather events. Although the research was motivated by the need to develop optimal salt application rates for parking lots and sidewalks, the results could be equally applicable for other transportation facilities after factors specific to the facility, such as traffic and dilution, are accounted for.

Prediction of Pavement Surface Temperature Using Meteorological Data for Optimal Winter Operations in Parking Lots

Pavement surface temperature is an essential requirement for the winter maintenance decision-making processes when determining suitable deicers and optimal application rates for transportation facilities such as parking lots. Currently, there are very few services/methods available to directly measure the pavement surface temperature. One method, the Road Weather Information System (RWIS), has less utility for parking lot and sidewalk maintenance purposes due to the operational constraints and high cost of using the RWIS service. This paper presents the results of a two-year field study focused on developing parametric and non-parametric models to forecast pavement surface temperature in parking lots based on a rigorous investigation of its correlation with common meteorological variables. The models were developed using available metrological data alongside measured pavement temperatures obtained through field observations. This study also evaluates the effects of different types of pavement (asphalt, interlocking bricks, and portland cement concrete) on pavement surface temperature. Evaluation results show that the developed models are able to predict pavement surface temperatures in different winter weather conditions with acceptable accuracy.