A Review of Urban Air Pollution Monitoring and Exposure Assessment Methods

Xingzhe Xie, Ivana Semanjski, Sidharta Gautama, Evaggelia Tsiligianni, Nikos Deligiannis, Raj Thilak Rajan, Frank J. Pasveer, Wilfried PhilipsISPRS Int. J. Geo Inf 2017 · DOI · OpenReview

TL;DR

This paper provides a comprehensive review of urban air pollution monitoring and exposure assessment methods, highlighting the shift from sparse fixed monitoring stations to denser mobile sensing. It synthesizes existing techniques and discusses the challenges and future directions for integrating these two data sources to achieve high-resolution, timely air quality assessments.

What Problem Does This Paper Address?

Urban air pollution poses significant risks to human health and the environment, necessitating accurate and high-resolution monitoring. Traditional methods rely on fixed monitoring stations, which are often sparsely distributed, leading to a lack of detailed spatial information. While mobile sensing technologies have emerged to address this sparsity, there's a need for a consolidated understanding of how to effectively acquire and analyze this new type of data, especially in conjunction with existing stationary data. This review aims to bridge this gap by examining various monitoring and assessment methods, with a particular focus on mobile sensing and the fusion of different data sources.

Methods

The paper categorizes air pollution assessment methods into three main groups:

  • Spatial Interpolation Approaches: These methods estimate pollutant concentrations at unmeasured locations based on data from surrounding monitoring stations. Techniques discussed include:
  • Spatial Averaging: Simple mean of nearby measurements.
  • Nearest Neighbor: Assigns the value of the closest station.
  • Inverse Distance Weighting (IDW): Assigns weights based on the inverse of the distance to the measurement points.
  • Kriging: A geostatistical method that considers spatial correlation among observations and provides estimates of uncertainty.
  • Data-Driven Spatial Prediction: Modern methods leveraging machine learning (e.g., neural networks) and compressed sensing to model spatial correlations and recover missing data.
  • Land-Use Regression (LUR) Models: These models establish relationships between pollutant concentrations and environmental characteristics of an area, typically derived from GIS data (e.g., traffic intensity, road length, land cover). They use multiple regression equations to predict pollutant levels.
  • Dispersion Models: These models simulate the physical and chemical processes of pollutant dispersion and transformation, predicting concentrations based on emission sources and environmental factors. Gaussian-based models (like CALINE and AERMOD) are commonly used, alongside non-Gaussian approaches.

The paper also discusses the application of these methods to mobile monitoring data, noting that researchers often adapt techniques developed for fixed stations. A significant challenge is the fusion of data from fixed and mobile sensors due to their differing spatio-temporal characteristics.

Datasets

The paper reviews numerous studies that utilize various datasets for air pollution monitoring and assessment. Specific datasets are not introduced as new benchmarks, but rather existing studies are referenced that employed data from:

  • Fixed Monitoring Stations: Data from national and local air quality monitoring networks across various cities and regions (e.g., Netherlands, USA, Canada, Belgium, China).
  • Mobile Monitoring Campaigns: Data collected using vehicles (vans, cars, buses, trams), bicycles, and even pedestrians equipped with sensors and GPS. These campaigns often focus on specific routes or areas of interest (e.g., traffic hotspots, urban canyons).

Key Insights and Ideas

  • The Rise of Mobile Sensing: The development of low-cost sensors and GPS technology has enabled mobile monitoring, significantly increasing the spatial density of air pollution measurements and offering a more granular view compared to sparse fixed stations.
  • Data Fusion is Key: Effectively combining data from fixed and mobile sensors is crucial for achieving comprehensive and accurate air quality assessments. This fusion presents significant challenges due to the inherent differences in data characteristics (e.g., temporal resolution, spatial coverage).
  • LUR Models' Versatility: Land-Use Regression models have proven to be a popular and effective tool for both stationary and mobile monitoring, capable of incorporating various environmental predictors to estimate pollutant concentrations.
  • Temporal vs. Spatial Resolution Trade-off: Mobile sensing excels at high spatial resolution but often at the cost of temporal resolution, especially in areas with low traffic or at night. Fixed stations offer consistent temporal data but lack spatial detail.
  • Pollutant-Specific Challenges: The effectiveness of different methods can vary depending on the pollutant being studied (e.g., CO vs. NO2 vs. PM).

Key Results

The paper is a review, so it doesn't present new experimental results. Instead, it synthesizes findings from numerous studies:

  • Spatial Interpolation: Studies show that Kriging generally provides more accurate estimates than IDW, especially when considering spatial correlations. However, the performance of interpolation methods can be highly dependent on the local monitor network density and the nature of the pollutant.
  • LUR Models: LUR models have demonstrated good predictive performance for mean annual concentrations of various pollutants (e.g., NO2, PM) in urban areas. They are often found to be superior to or equivalent to geo-statistical methods for exposure assessment.
  • Mobile Sensing Applications: Mobile monitoring has been successfully applied to estimate traffic-related pollution, personal exposure, and to create high-resolution air quality maps.
  • Fusion Efforts: While research on fusing mobile and fixed sensor data is still in its early stages, initial efforts suggest potential for improved accuracy and coverage.

Limitations and Future Work

  • Limitations of Mobile Data: Mobile sensing campaigns often cover specific routes or time periods, leading to uneven spatial and temporal coverage. This can result in limited data in low-traffic conditions or at night, impacting long-term estimation.
  • Challenges in Data Fusion: The inherent differences in spatio-temporal characteristics between fixed and mobile sensor data pose a significant challenge for effective fusion. Existing research often neglects temporal variations and the impact of sensor velocity.
  • Transferability: Ensuring the transferability of findings and models across different regions and campaigns remains a challenge.
  • Future Work:
    • Developing strategies to handle temporal variations between different locations (e.g., sensor velocity).
  • Addressing reduced and inconsistent temporal resolution in mobile sensing campaigns.
  • Developing robust strategies for longitudinal air pollution monitoring and the fusion of data from fixed and mobile networks.
  • Further research into combining mobile and fixed pollution data is needed.
  • Developing standardized air quality indicators to facilitate comparability across different initiatives and regions.

Summary updated

Abstract

The impact of urban air pollution on the environments and human health has drawn increasing concerns from researchers, policymakers and citizens. To reduce the negative health impact, it is of great importance to measure the air pollution at high spatial resolution in a timely manner. Traditionally, air pollution is measured using dedicated instruments at fixed monitoring stations, which are placed sparsely in urban areas. With the development of low-cost micro-scale sensing technology in the last decade, portable sensing devices installed on mobile campaigns have been increasingly used for air pollution monitoring, especially for traffic-related pollution monitoring. In the past, some reviews have been done about air pollution exposure models using monitoring data obtained from fixed stations, but no review about mobile sensing for air pollution has been undertaken. This article is a comprehensive review of the recent development in air pollution monitoring, including both the pollution data acquisition and the pollution assessment methods. Unlike the existing reviews on air pollution assessment, this paper not only introduces the models that researchers applied on the data collected from stationary stations, but also presents the efforts of applying these models on the mobile sensing data and discusses the future research of fusing the stationary and mobile sensing data.
BibTeX
@article{xie2017review,
  author = {Xingzhe Xie and Ivana Semanjski and Sidharta Gautama and Evaggelia Tsiligianni and Nikos Deligiannis and Raj Thilak Rajan and Frank J. Pasveer and Wilfried Philips},
  title = {{A Review of Urban Air Pollution Monitoring and Exposure Assessment Methods}},
  journal = {ISPRS Int. J. Geo Inf.},
  year = {2017},
  doi = {10.3390/ijgi6120389},
}
Figure