Analyzing the Effects of Road Type and Rainy Weather on Fuel Consumption and Emissions: A Mesoscopic Model Based on Big Traffic Data

Road transportation accounts for significant percentages of urban energy consumption and carbon emissions. Therefore, it is important to predict and analyze the fuel consumption and emissions for on-road vehicles, which are varied under different conditions. Previous studies have shown that some traffic elements such as road type and weather condition have considerable influence on transportation fuel consumption and emissions. However, limited to the data availability, most of the existing studies focus on specific routes or scenarios, and few of them consider the effects of road type and weather condition systematically at large scale. In this research, a data-driven mesoscopic model was developed to investigate the effects of road type and weather condition on the link-level fuel consumption and emission factors based on big traffic data. This built model utilized the neural network for the prediction algorithm with inputs including road type, weather condition, and link-level aggregated operation data obtained through link-based segregation over trajectory snippets. The investigation was carried out with real-world big traffic data collected from 10,944 taxis over a 2-month period of operation in Shenzhen, and produced reliable predictions for four road types with clear and rainy weather conditions. Both statistical analysis and model prediction results showed that fuel consumption and emission factors are lower in low-speed range for freeway and expressway, and are lower in middle-speed range for main road and secondary road. In addition, rainy weather condition tends to have lower fuel consumption and emission factors than clear weather condition.