Empirical Studies on Traffic Flow in Inclement Weather

1.0 Introduction

1.1 Background

Weather conditions affect the operation of the national transportation system by changing the driving environment as well as the behavior of drivers, who modify their individual headways, target speeds, or other travel parameters in reaction to specific weather events. The individual reactions of these drivers to weather, in turn, directly impact overall system performance. From rain showers in Honolulu, to fog in San Francisco, to snow in Minneapolis, to storms in Boston, local transportation system managers understand the necessity of planning for the consequences of diverse weather conditions on a daily basis. However, at a national scale, universally applicable tools to understand and mitigate the effects of weather on transportation systems remain in short supply.

Weather causes a variety of impacts on transportation systems during and after weather events. These impacts can be short- or long-term, and direct or indirect. However, the linkages between inclement weather conditions and traffic flow in existing analysis tools remain tenuous. Transportation managers require information on the impacts of weather on traffic in order to apply advisory, control, and treatment strategies on their systems to deal with weather events. As a first step in strengthening this analysis capability, researchers and practitioners must have access to weather data sets and analytical models that can connect traffic flow to specific weather impacts. Without this information, managers will continue to rely on heuristic, local methods for determining impacts. This approach confounds coordination between managers and confuses the national approach to weather-related traffic flow impact analysis by failing to provide an acceptable knowledge base and techniques.

Oak Ridge National Laboratory conducted a study that estimated the delay experienced by American drivers in 1999 at 46 million hours on major United States highways due to adverse weather conditions defined for this study as fog, ice, and snowstorms (Han, et al., 2003). These conditions covered 65 percent of United States territory and impacted 68 percent of the nation’s population. Researchers estimated that the occurrence of inclement weather increased overall travel times by approximately 7 to 36 percent. The weather events studied are highly seasonal events and occur most frequently during the winter months. Approximately two-thirds of traffic flow and delay impacts during the winter season can be attributed to adverse weather.

While severe winter storms, hurricanes, or excessive flooding can result in major stoppages or evacuations of transportation systems or evacuations and cost millions of dollars, day-to-day weather events such as rain, fog, snow, and freezing rain can have serious impacts on roadway mobility and the safety of the transportation system users. These weather events can result in increased fuel consumption, delay, number of accidents, and significantly impact the performance of the transportation system.

1.2 Motivation for This Study

Transportation managers strive to improve safety and mobility within their systems. Currently, weather events add an element of variability to their day-to-day operations. However, these events can be forecast, and their impacts predicted, if data and decision support tools are available and utilized effectively. Improved predictive capabilities are critical to effective operational response. Tailored road weather observations and forecasts that provide information on the location, characteristics, and duration of weather events enable transportation managers to assess the scope, severity, and overall impacts of weather events on their systems.

Many municipalities have recently established Road Weather Information Systems (RWIS), a combination of technologies that use historic climate and current road weather data to develop information that aid roadway-related decisions. These systems improve the weather forecasts available to transportation managers and provide a wealth of information for analysis by providing weather and pavement condition data. While sensor station weather data provides benefit in reporting real-time conditions to managers, it has the potential to provide more value if archived and used to evaluate traffic management strategies.

Tools that use historic traffic data and weather forecast data to predict traffic flow conditions are important to help identify where and how weather events will impact traffic conditions. These can also enable managers to determine areas and conditions that are most vulnerable to incidents, and estimate the safety impacts of both weather events and response strategies. Long-term planning tools facilitate development and evaluation of weather-responsive traffic management strategies. Models that predict short-term traffic flow using historical and real-time data are useful in assessing its current impacts on the transportation system and determining implementation of a specific strategy. With these tools, transportation managers can implement effective advisory strategies that provide information on weather and traffic conditions to the public and control strategies that alter the state of roadway devices to regulate traffic flow and roadway capacity during a weather event by modifying signal timings/plans and speed limits, closing roads, and implementing detour routes.

The aim of this research is to investigate and quantify the impacts of various weather events on traffic flow in order to provide transportation managers with information and tools they can utilize to improve mobility and safety on the transportation network.

1.3 Scope of Report

The early phases of this project included background research on the impact of weather on traffic flow. The first activity, a literature review, is documented in Section 2. This was followed by a search of available data sources for both weather and traffic information. The results of this effort are summarized in Section 3. The research plan and models proposed for analysis are described in Section 4 and the research findings are detailed in Section 5. Section 6 contains recommendations for further research.