Impacts Of Rainfalls And Snowfalls On Urban Roads Traffic Performance

The rapid development of cities and the explosive increase of vehicle population necessitate a higher operational efficiency of urban road networks. As a result, the research on the impacts of adverse weathers, such as rainfalls and snowfalls, on traffic operations is becoming particularly important. An extensive study of the impacts of rainfalls and snowfalls on urban traffic is of practically important for improving the corresponding management measures and ensuring a high-efficiency of traffic performance on urban roads. In this context, this thesis is intended to analyze the impacts of rainfalls and snowfalls on the traffic operations on urban roads for different time periods, congestion levels and precipitation intensities, striving to develop the relationships between the intensity of the rainfall or snowfall and the variation of the speed, traffic performance index (TPI), and road capacity under different conditions.First, the thesis conducts the literature review, from perspectives of:(1) the impact analysis of rainfalls and snowfalls on the traffic performance, and (2) the methods of traffic data quality control. It analyses the influential factors for the impact studies of rainfalls and snowfalls on traffic operations, upon which a State Preference (SP) survey is designed for analyzing the impacts of rainfalls and snowfalls on traffic operations. Based on the analysis result of the SP survey, the research directions and methods for the impact analysis of rainfalls and snowfalls are determined. Further, the analysis methods of the characteristics and availability of the weather and traffic data are designed. Subsequently, the thesis analyses the characteristics of the precipitation, network speed, and traffic performance index (TPI) in Beijing according to the weather and traffic data available in Beijing and the result of the SP survey, and further analyses the availability of the data. With the results of the characteristics and availability analysis, the thesis designs the methods of data matching and processing, based on a GIS platform and the temporal and spatial characteristics of the weather and traffic data. Afterward, the weather and traffic data are processed from the perspectives of the road class, time period, congestion level, and precipitation intensity.Second, the thesis conducts the impact analysis of the rainfalls on urban road traffic operations for different time periods and congestion levels based on the results of the data availability analysis and the data matching and processing. The following results are obtained:under the same conditions of the precipitation intensity, time period and congestion level, speeds on the expressways are affected the most by the rainfall, followed by the major arterials and the local streets. Among different time periods, the precipitation affects the speed and TPI the most during peak hours, while the least at night. During peak hours, the travel speeds on the expressways, major arterials, minor arterials and local streets have decreased by 15%-19%,12%-15% and 9%-12% respectively. Among different congestion levels, there is a minimum impact by the rainfall on travel speeds when the traffic congestion level is at the "free-flow." The "heavy congestion" is the most sensitive traffic condition to rainfalls, in which the speeds are reduced by 25-30%,20%, and 15% for expressways, major arterials, and minor arterials and local streets respectively under the light and moderate rains. Further, it is found through the parameter calibration of the Van Aerde traffic flow model that the road capacities are reduced by 4%,6%, and 9% for the 2nd,3rd, and 4th Ring Roads respectively under rainfalls. Afterward, the thesis develops the speed prediction correction model under the rainfalls for different road classes and time periods.Third, the snowfall data is insufficient to support the analysis similar to that for rainfalls based on the snowfall intensity, time period, and speed correction, because of the limitation of the sample size and data quality of snowfalls in Beijing. It is observed that the time periods of the snowfall have a different influence on the traffic from those of the rainfall, thus, the thesis conducts the impact analysis of snowfalls, from the perspective of snowfall time periods for the scenarios of "snow begins the day before the travel" and "snow begins on the day of the travel." Further, the snowfall impacts are analyzed quantitatively by categorizing two periods of snowfalls, the beginning period and the later period of snowfalls. During the peak hours, the travel speeds on expressways, major arterials and local streets have decreased by 10.3 km/h (24%),4.2 km/h (16%), and 3.1 km/h (13%) respectively at the beginning period of snowfalls, and increased by 10.3 km/h (28%),5.3 km/h (24%), and 3.4 km/h (16%) respectively at the later period of snowfalls.Finally, the thesis summarizes the findings of this study and analyzes its differences from existing studies based on the above analysis of the impacts of rainfalls and snowfalls on traffic operations. Recommendations on future study are presented.