Air Pollution Hotspot Detection and Source Feature Analysis using Cross-Domain Urban Data

Air pollution is a major global environmental health threat, in particular\nfor people who live or work near pollution sources. Areas adjacent to pollution\nsources often have high ambient pollution concentrations, and those areas are\ncommonly referred to as air pollution hotspots. Detecting and characterizing\npollution hotspots are of great importance for air quality management, but are\nchallenging due to the high spatial and temporal variability of air pollutants.\nIn this work, we explore the use of mobile sensing data (i.e., air quality\nsensors installed on vehicles) to detect pollution hotspots. One major\nchallenge with mobile sensing data is uneven sampling, i.e., data collection\ncan vary by both space and time. To address this challenge, we propose a\ntwo-step approach to detect hotspots from mobile sensing data, which includes\nlocal spike detection and sample-weighted clustering. Essentially, this\napproach tackles the uneven sampling issue by weighting samples based on their\nspatial frequency and temporal hit rate, so as to identify robust and\npersistent hotspots. To contextualize the hotspots and discover potential\npollution source characteristics, we explore a variety of cross-domain urban\ndata and extract features from them. As a soft-validation of the extracted\nfeatures, we build hotspot inference models for cities with and without mobile\nsensing data. Evaluation results using real-world mobile sensing air quality\ndata as well as cross-domain urban data demonstrate the effectiveness of our\napproach in detecting and inferring pollution hotspots. Furthermore, the\nempirical analysis of hotspots and source features yields useful insights\nregarding neighborhood pollution sources.\n