Influence of three-dimensional patterns of buildings on urban atmospheric pollutants in metropolises of China

• An integrated analysis involving 2-D and 3-D building patterns on urban air pollutant. • SEM and BRT statistical analysis was an efficient quantitative cross validation method. • Subregions were divided considering Chinese climate, terrain regions and seasons. • The 3-D building patterns has greater impact on air pollutant in low concentration area. The atmospheric environment affects the residents’ health and welfare, especially in urban areas, which are severely influenced by ground ventilation conditions determined by land use and urban morphology. Quantitatively analyzing the relationship between urban landscape patterns and atmospheric pollution remains a challenge. In our study, the relationship between urban three-dimensional(3-D) patterns on four atmospheric pollutions(SO 2 , NO 2 , PM 2.5 and PM 10 ) at national, gradient terrain and heating/non-heating in winter subregions were analyzed based on the cross-validation of boosted regression tree and structural equation modeling for reducing research uncertainty. Cities in central and northern China exhibited highest concentrations and these pollutants were significantly influenced by two-dimensional(2-D) patterns at the national subregion, including impervious ratio. In heating regions, the main factors influencing these pollutants were related to 2-D pattern, such as impervious ratio, catering quantity, and road density. In non-heating regions, 3-D patterns, including floor area ratio and high building ratio, were frontal factors. In the terrain gradient, 2-D landscape patterns significantly influenced atmospheric pollutants, with patch density, catering quantity, and number of buildings in second terrain step and impervious ratio in third terrain step. 2-D and 3-D landscape patterns significantly impact atmospheric pollutant distribution. Nonetheless, in low concentration regions, 3-D patterns have greater impact on atmospheric pollutant concentrations due to their effects on urban form and microclimate. In the context of continuous air quality control and low concentrations in urban atmosphere, these multi-scale analysis results enrich research on atmospheric pollution and urban morphology, offering valuable insights for urban planning and atmospheric pollution control.