Comparison of residential and mobility-integrated air pollution exposures from tracking campaigns and agent-based modelling in Switzerland and the Netherlands – Nature

Report on Comparison of Residential and Mobility-Integrated Air Pollution Exposures in Switzerland and the Netherlands

Introduction

Air pollution is a significant environmental health risk, contributing to millions of deaths globally each year. Long-term exposure to pollutants such as nitrogen dioxide (NO₂) and fine particulate matter (PM_2.5) has been linked to adverse health outcomes including increased mortality. Epidemiological studies typically estimate exposure based on outdoor air pollution levels at residential addresses, potentially overlooking individual mobility and time-activity patterns. This may lead to exposure misclassification, affecting the precision and validity of health effect assessments.

This report emphasizes the alignment of such exposure assessments with the United Nations Sustainable Development Goals (SDGs), particularly SDG 3 (Good Health and Well-being), SDG 11 (Sustainable Cities and Communities), and SDG 13 (Climate Action), by improving understanding of air pollution exposure and its health impacts.

Objectives

1. To evaluate the differences between residential address-based air pollution exposure estimates and mobility-integrated exposures derived from GPS tracking and agent-based modeling (ABM).
2. To assess the applicability of these methods in two European countries with differing commuting patterns: Switzerland and the Netherlands.

Materials and Methods

Tracking Campaign

Two tracking campaigns were conducted in 2022/23 involving 686 participants (489 in Basel region, Switzerland, and 189 in the Netherlands). Participants were monitored over two weeks using GPS trackers and a mobile app to record location and time-activity diaries. Baseline questionnaires collected demographic, health, and work location data.

Agent-Based Modeling (ABM)

ABM simulated individual mobility and activities based on demographic profiles and national travel survey data. The model generated multiple realizations of daily activities including commuting, work, shopping, and recreation, assigning air pollution exposures by overlaying simulated routes with hourly pollutant concentration surfaces. ABM profiles included residential, homemaker, and commuter categories, reflecting diverse population segments.

Exposure Estimation

• Residential exposures were assigned based on annual average pollutant concentrations at participants’ home addresses.
• Mobility-integrated exposures were calculated using GPS tracking data and ABM simulations, incorporating temporal and spatial variations in NO₂ and PM_2.5 concentrations.
• Exposure surfaces were derived from high-resolution land-use regression models and monitoring data for both countries.

Statistical Analysis

Comparisons between residential, GPS-based, and ABM-based exposures were conducted using Bland-Altman plots, scatterplots, and coefficients of determination (R²). Correlation strengths were categorized from weak to very strong. Sub-analyses included ABM with known workplace locations and simulations using single versus mean realizations.

Results

Participant Demographics

• Participants were predominantly female, employed, and from urbanized areas with higher education and income levels compared to national populations.
• Exposure levels in tracking campaigns reflected urban settings, with higher mean NO₂ and PM_2.5 concentrations than broader cohorts in Switzerland; Dutch exposures were comparable to national cohorts.

Exposure Comparisons

1. Residential vs. GPS Tracking: Strong correlations for NO₂ (R² > 0.76 in Switzerland and 0.79 in the Netherlands) and moderate for PM_2.5 (R² = 0.56 in the Netherlands).
2. GPS Tracking vs. ABM: Strong agreement for NO₂ (R² > 0.77) and variable for PM_2.5 (stronger in Switzerland, R² = 0.80, than in the Netherlands, R² = 0.54).
3. Residential vs. ABM: Highest correlations observed (R² > 0.96 for both pollutants), indicating that ABM effectively simulates mobility-integrated exposures.

Impact of Known Workplace Locations

In subsets with known work addresses, ABM incorporating actual workplace data showed very strong correlations with residential exposures (R² between 0.81 and 0.90), supporting the robustness of residential-based exposure estimates.

Simulation Variability

Using the mean of 50 ABM realizations yielded higher correlations with residential exposures compared to single random realizations, indicating the importance of accounting for variability in mobility patterns.

Discussion

Key Findings

• Residential address-based air pollution exposure estimates correlate strongly with mobility-integrated exposures derived from GPS tracking and ABM.
• ABM provides a scalable and representative method for estimating long-term air pollution exposures in large populations, aligning with SDG 11 by supporting sustainable urban health assessments.
• Inclusion of workplace location data enhances exposure assessment accuracy, emphasizing the value of comprehensive data collection in epidemiological studies.
• Mobility patterns can influence individual exposure levels, particularly for those residing in low pollution areas who commute to higher pollution zones, highlighting the need for integrated exposure assessments.

Comparison with Previous Research

The findings are consistent with prior studies demonstrating high correlations between residential and mobility-integrated exposures. This supports the continued use of residential address-based exposure assessments in large-scale epidemiological research, contributing to SDG 3 by enabling effective public health interventions.

Strengths and Limitations

• Strengths include the large sample size, dual-country design, and integration of empirical and modeling approaches.
• Limitations involve non-representative participant demographics and urban-centric study settings, suggesting further research is needed in rural contexts.
• The study focuses on long-term exposures to NO₂ and PM_2.5; applicability to other pollutants requires additional investigation.

Conclusions

This study demonstrates that residential address-based air pollution exposure assessments provide reliable estimates comparable to mobility-integrated methods using GPS tracking and ABM. These findings support the use of residential exposures in epidemiological studies on long-term health effects of air pollution, facilitating progress towards SDG 3 (Good Health and Well-being) and SDG 11 (Sustainable Cities and Communities). Incorporating mobility data through ABM, especially when workplace locations are known, offers a valuable enhancement for exposure assessment, aligning with SDG 13 (Climate Action) by improving understanding of pollution exposure dynamics.

Implications for Sustainable Development Goals (SDGs)

• SDG 3: Good Health and Well-being – Improved exposure assessment methods enhance epidemiological studies, informing policies to reduce air pollution-related health burdens.
• SDG 11: Sustainable Cities and Communities – Understanding mobility patterns and pollution exposure supports urban planning for healthier environments.
• SDG 13: Climate Action – Accurate exposure data aids in evaluating the effectiveness of air quality interventions and climate policies.

1. Sustainable Development Goals (SDGs) Addressed in the Article

1. SDG 3: Good Health and Well-being
   • The article focuses on the health effects of long-term exposure to air pollution, which is directly related to ensuring healthy lives and promoting well-being for all ages.
2. SDG 11: Sustainable Cities and Communities
   • The study involves urban and regional air pollution exposure assessments in Switzerland and the Netherlands, addressing sustainable urban environments.
3. SDG 13: Climate Action
   • Although not explicitly stated, air pollution control is closely linked to climate action, as reducing pollutants like NO₂ and PM_2.5 contributes to climate mitigation.

2. Specific Targets Under the Identified SDGs

1. SDG 3: Good Health and Well-being
   • Target 3.9: By 2030, substantially reduce the number of deaths and illnesses from hazardous chemicals and air, water, and soil pollution and contamination.
2. SDG 11: Sustainable Cities and Communities
   • Target 11.6: By 2030, reduce the adverse per capita environmental impact of cities, including by paying special attention to air quality and municipal and other waste management.
3. SDG 13: Climate Action
   • Target 13.2: Integrate climate change measures into national policies, strategies, and planning, which includes actions to reduce air pollution emissions.

3. Indicators Mentioned or Implied in the Article to Measure Progress

1. Air Pollution Concentration Indicators
   • Annual average concentrations of NO₂ (Nitrogen Dioxide) and PM_2.5 (fine particulate matter) are used as key indicators to assess exposure levels.
   • Hourly average concentrations and spatial-temporal distribution of these pollutants are also considered to capture exposure variations.
2. Exposure Assessment Metrics
   • Residential-based exposure versus mobility-integrated exposure (using GPS tracking and Agent-Based Modeling) are compared to evaluate accuracy and bias in exposure assessment.
   • Correlation coefficients (R²) between different exposure assessment methods serve as indicators of agreement and reliability.
3. Health Outcome Indicators (Implied)
   • Though not directly measured in this article, the study references mortality and morbidity related to long-term air pollution exposure, implying the use of health statistics as indicators in related epidemiological studies.

4. Table of SDGs, Targets, and Indicators Relevant to the Article

Source: nature.com