Abstract:Urban air pollution is one of the serious environmental problems in the development of high-density cities, especially for traffic-related pollution. Effective urban environmental monitoring technologies are helpful to predict and control the air pollution emitted by traffic sectors. However, the complex temporal and spatial characteristics of air pollution in urban areas pose a great challenge for its monitoring, prediction, and control. This work investigated on the spatiotemporal air pollution patterns and its correlation mechanism with urban elements (such as traffic) based on limited data of monitoring stations, which is significant for the accurate prediction and regulation of urban environment. Regional Cluster Monitoring (RCM) methods were proposed through integrating multi-source data, including data from U.S. Geological Survey (USGS) Landsat, OpenStreetMap (OSM) road network , and environmental monitoring stations to categorize different monitoring stations. According to different urban areas (defined as U: 'urban' or S: 'suburban') and traffic densities (T: 'traffic' or B: 'background'), monitoring stations areas were categorized into U-B, U-T, S-B, and S-T. Then the spatiotemporal characteristics of traffic-related pollutants in different region categories were analyzed. The rapid prediction of air pollution distribution in different categories of urban regions was realized by RCM method-based Long Short-Term Memory (LSTM) neural network. Results showed that the proposed methods could accurately represent the spatiotemporal distribution characteristics of urban air pollution. The mean absolute percentage errors (MAPE) between the predicted results and actual observed values were 20.31% to 28.05% (absolute errors were 3.9-14.3 μg/m3). This work can significantly impact on low-cost & effective monitoring and control strategies for urban area, and further contribute to comprehend the spatial-temporal distribution patterns of urban traffic-related air pollution from limited data of monitoring stations.

Dynamic Traffic Data in Machine-Learning Air Quality Mapping Improves Environmental Justice Assessment

High-resolution mapping of regional traffic emissions using land-use machine learning models

Urban Air Pollution Mapping Using Fleet Vehicles As Mobile Monitors and Machine Learning

Examining air pollution exposure dynamics in disadvantaged communities through high-resolution mapping

Forecasting Smog-Related Health Hazard Based on Social Media and Physical Sensor.

Data-Driven Air Quality Characterization for Urban Environments: A Case Study

Deep-MAPS: Machine Learning based Mobile Air Pollution Sensing

Developing high-resolution PM2.5 exposure models by integrating low-cost sensors, automated machine learning, and big human mobility data

Combining Google traffic map with deep learning model to predict street-level traffic-related air pollutants in a complex urban environment

Fine-Scale Spatiotemporal Air Pollution Analysis Using Mobile Monitors on Google Street View Vehicles

Fine-Grained Air Pollution Inference with Mobile Sensing Systems: A Weather-Related Deep Autoencoder Model

Machine Learning for Urban Air Quality Analytics: A Survey

Deep-AIR: A Hybrid CNN-LSTM Framework for Air Quality Modeling in Metropolitan Cities

Weather and the City: Machine Learning for Predicting and Attributing Fine Scale Air Quality to Meteorological and Urban Determinants

Temporal-Spatial Aggregated Urban Air Quality Inference With Heterogeneous Big Data

A Machine-Learning-Aided Visual Analysis Workflow for Investigating Air Pollution Data

Spatial and temporal urban air pollution patterns based on limited data of monitoring stations

Novel Approach for Predicting the Air Quality Index of Megacities through Attention-Enhanced Deep Multitask Spatiotemporal Learning

Data-driven Air Quality Characterisation for Urban Environments: a Case Study

Modeling fine-grained spatio-temporal pollution maps with low-cost sensors