ITS - Intelligent Transportation Systems Report ITS Home Page
Table of Contents

Chapter 1. Introduction

The disruptive effect of inclement weather on traffic is well known to most drivers and travelers, and is a challenging issue to traffic engineers and managers. In addition to its staggering impact on safety (it is estimated that about 28% of all highway crashes and 19% of all fatalities involve weather-related adverse road conditions as a factor), adverse weather results in reduced service capacity (often at the most critical of times), diminished reliability of travel, reflected in considerable variability and unpredictability, and greater risk of accident involvement. It is well documented that weather exerts significant impact on several key traffic flow parameters, such as free flow speed and capacity (e.g. Kockelman, 1998; Smith et al. (2004) and a recent FHWA report (Hranac et al., 2006) summarizing empirical studies on traffic flow in inclement weather). In addition, adverse weather often affects tripmaker decisions of travel mode, route, timing, destination, or even whether to make the trip at all (e.g. telecommute or teleshop instead). Thus, weather affects both the supply and demand sides of transportation. Recognizing it into transportation operations and management has the potential to improve the performance of the transportation system at times where such improvement is most critically needed.

Yet, an assessment of current and past practice, as well as of the major reference documents typically used by practicing traffic engineers, quickly reveals that there is little out there to guide traffic planners and engineers in dealing with adverse weather on a regular basis. There appears to be a perception that there is not much that one could do to deal with such situations, other than caution drivers to stay home, drive slowly, or be more alert. This perception may be rooted in three inter-related causes: (1) absence of specific actions and measures, and accepted conditions for their application, that could be deployed specifically to manage traffic under adverse weather; (2) lack of tools to support such management decisions, including analysis/evaluation of the impact of contemplated actions and design of interventions, both off-line and on-line; and (3) insufficient understanding of the [qualitative and quantitative] effects of adverse weather of varying characteristics on traffic flow, on the performance of different types of facilities with varying geometric and operational features, and on the response of users to the weather phenomena as well as to contemplated control and management actions.

Advances in sensor technologies and continuing deployment of intelligent transportation system (ITS) architectures provide an important opportunity to anticipate, mitigate, and intervene through various advisory and control measures to improve traffic conditions in the presence of inclement weather. The premise of ITS is the ability to sense prevailing conditions, anticipate unfolding future conditions, and rapidly devise actions to optimize system performance in real-time. Dealing with adverse weather requires not only sensing of traffic conditions, but also the ability to forecast the weather in real-time for operational purposes. Recognizing the importance of tying weather and traffic management together in areas exposed to extreme weather situations, such as hurricanes and floods, some Traffic Management Centers (TMC) co-locate the weather service personnel with the usual traffic management agencies (police, traffic operators, Emergency Medical Services). Another relevant initiative is the Clarus weather data system, intended to provide traffic management centers with accurate real-time weather information (Pisano and Goodwin, 2002; Mixon-Hill Inc. et al., 2005; Pisano, Alfelor, et al., 2005; FHWA Clarus web site, at http://www.its.dot.gov/clarus/index.htm). The weather information, along with the traffic information obtained from ITS sensors, enable promising new opportunities to improve traffic operations and management under inclement weather conditions. In addition, the Clarus system will eventually be coordinated with IntelliDriveSM systems such that both vehicular information as well as weather data can be obtained for traffic management.

A critical methodological capability in the above architecture is a Traffic Estimation and Prediction System (TrEPS). Because the dynamics of traffic systems are complex, many situations call for strategies that anticipate unfolding conditions instead of adopting a purely reactive approach. Real-time simulation of the traffic network forms the basis of a state prediction capability that fuses historical data with sensor information, and uses a description of how traffic behaves in networks to predict future conditions, and accordingly develop control measures. The estimated state of the network and predicted future states, in terms of flows, travel times, and other time-varying performance characteristics, are used in the on-line generation and real-time evaluation of a wide range of measures, including information supply to users, VMS displays, coordinated signal timing for diversion paths, as well as weather-related interventions (through variable speed displays, advisory information, signal timing adjustments and so on). The core of the descriptive DTA capability is a traffic simulation model, intended to capture the dynamics of traffic flow movement in the network (Jayakrishnan et al. 1994; Mahmassani 1998, 2001).

Recognizing the need for prediction in advanced traffic management systems, the FHWA funded R&D into the methodological foundations of simulation-based DTA for TrEPS application, and subsequently supported the development of two prototypes, DYNASMART-X and DynaMIT, both of which adopted similar methodological decisions with regard to the underlying simulation logic-specifically, to use a mesoscopic approach in which individual particles (vehicles) move according to local speeds determined consistently with (macroscopic) relations among averages of speed and density. This mesoscopic approach for network-level TrEPS defines the state of the art in this domain. However, these tools have to date only been calibrated and tested under "normal" weather conditions. In other words, no provision has been made to explicitly capture the behavioral phenomena that determine traffic patterns under adverse weather, predict how traffic might be impacted by such weather, and how it might respond to various advisory and regulatory interventions aimed at managing traffic during such conditions. Therefore, while a major need for on-line estimation and prediction arises precisely because of unanticipated weather perturbations, the tools developed for such applications did not initially have the ability to represent traffic behavior under such conditions, or in response to the possible interventions.

To address the above-mentioned deficiency this project is aimed at developing weather-sensitive traffic prediction and estimation models and incorporate them in existing traffic estimation and prediction systems. In particular, the following tasks are performed to achieve these goals:

The remainder of this report is organized as follows. A literature review is presented in Chapter 2, which covers two major aspects: (1) impact of weather events on traffic system performance (supply side); (2) traveler behavioral responses to weather events and related traffic advisory information (demand side). This is followed by a review of existing traffic prediction/estimation models and systems in Chapter 3, which can be used for both planning and real-time traffic management applications at the corridor and network levels (i.e., DYNASMART, DynaMIT). Chapter 4 presents a methodology to model the weather impact in DYNASMART. The principal supply-side and demand-side elements that would be affected by adverse weather are identified and modeled in the framework of traffic estimation and prediction systems. These proposed weather-related features are implemented in DYNASMART, as described in Chapter 5. In Chapter 6, calibration procedures as well as the results are presented. Chapter 7 demonstrates the application of the weather-sensitive traffic estimation and prediction model to a real world network, focusing on two aspects: (1) assessing the impacts of adverse weather on transportation networks; and (2) evaluating effectiveness of weather-related advisory/control strategies in alleviating traffic congestion due to adverse weather conditions. Finally, Chapter 8 concludes the project and discusses further research directions.

Previous | Next