Quantifying spatially varying impacts of public transport on NO _2 concentrations with big geo-data

nthropogenic NO _2 concentrations cause climate change and human health issues. Previous studies have focused on the contribution of traffic factors to NO _2 emissions but have ignored the spatially varying impact of public transport supply and demand on high-resolution NO _2 concentrations. This study first applies a two-stage interpolation model to generate a high-resolution urban NO _2 concentration map originating from satellite measurement products. Then, we formulate 12 explanatory indicators derived from a fusion of massive big geo-data including smart card data and point of interest information, to represent the specific degree of public transport supply and citizens’ demand. Furthermore, a geographically weighted regression is applied to quantify the spatial variation in the effect of these indicators on the urban NO _2 concentrations. The result shows that public transportation coverage, frequency, and capabilities as public transport supply indicators in metropolitan and suburban areas have a two-way influence on the NO _2 emissions. However, among public transport demand indicators, the economic level has a significant positive impact in most areas. Our findings can provide policy implications for public transportation system optimization and air quality improvement.