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:

Forecasting of Road Surface Temperature Using Time Series, Artificial Neural Networks and Linear Regression Models

Abstract:

This research investigates the feasibility of applying simple statistical models for forecasting road surface temperature at locations where RWIS data are available. Three commonly used modeling techniques are considered and those are time-series analysis, linear regression and artificial neural networks (ANN). A data set from a RWIS station is used for model calibration and validation. This paper describes the major findings with a specific focus on the generalization capability of the models. The analysis indicates that multi-variable and ANN are the most competitive technique with lowest forecasting errors.

Source(s):

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

Date: 2007

Author:

Fu, Hashemloo, Fu, Kim

Keywords:


Pavement temperature
Road weather information system (RWIS)
Environmental Sensor Station (ESS)
Observing network

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