Dataset for "Revealing nonlinear associations between built-environment factors and pedestrian-weighted air pollution exposure using GeoXAI"

This repository contains the processed dataset supporting the findings reported in the manuscript: Revealing nonlinear associations between built-environment factors and pedestrian-weighted air pollution exposure using GeoXAI The dataset was developed for a street-scale assessment of pedestrian-weighted air pollution exposure (PWEAP) in the core area of Honggutan District, Nanchang, China. The repository contains processed data only. Raw mobile monitoring records and third-party proprietary datasets are not redistributed where licensing restrictions apply. The processed dataset includes: Street-segment IDs and spatial coordinates Background-normalized street-level PM2.5 concentrations Pedestrian activity intensity derived from Baidu WiseEye Pedestrian-weighted exposure to air pollution (PWEAP) Nineteen built-environment variables, including traffic indicators, road-network characteristics, POIs, NDVI, NDWI, building morphology and housing price Three temporal scenarios: Weekday Peak (WP) Weekday Off-Peak (WOP) Weekend (WE) The study contains 1,972 street segments within the study area.