Road Weather Management Program
photos of lightning, trucks plowing snow, an empty road before a storm, and an evacuation
Office of Operations 21st century operations using 21st century technologies

Best Practices for Road Weather Management Version 2.0

Best Practices > Publications > Publication Titles

Title:

Probabilistic Models for Discriminating Road Surface Conditions Based on Friction Measurements

Abstract:

This paper presents two statistical models for discriminating different types of road surface contaminants based on friction measurements and other road condition data. The first model is a disaggregate logit model which can be used to predict the probability that a road surface is covered by snow or in bare condition based on direct friction measurements and other available road weather data. The second model is a aggregate logit regression model that uses aggregated measures over a section of road as input to distinguish two sub snow cover states, namely, full snow cover and partial snow cover. The proposed models are calibrated using field data collected from a maintenance route in Ontario, Canada and are shown high discrimination power based on holdout data sets.

Source(s):

87th Transportation Research Board (TRB) Annual Meeting, University of Waterloo and Ontario Ministry of Transportation (Canada). For an electronic copy of this resource, please direct your request to WeatherFeedback@dot.gov.

Date: 2007

Author:

Fu, Feng, Perchanok

Keywords:


Pavement friction
Forecast/Prediction
Snow

PDF files can be viewed with the Acrobat® Reader®.
Office of Operations