Bridging gap in air pollution health assessment: Integrating Earth Observation, mobility data and spatial analysis for air pollution health risk analyses.

Kurzfassung

When performing a health risk assessment related to air pollution by means Earth Observation (EO) data two of the main challenges are: (1) the challenge of accurately assessing population exposure and (2) the inconsistent spatial and temporal coverage of air pollution data, especially in remote areas. Health risk is determined by the interaction of three components: hazard, exposure, and vulnerability. When investigating the health risk from outdoor air pollution, exposure at individual level is oftentimes of challenging quantification. Therefore, epidemiological studies are typically conducted using static residential data, neglecting the dynamic of human mobility and its impact on exposure. In recent years, the popularity of epidemiological studies exploiting an ecological approach is rising due to the increasing availability of publicly available of geospatial data. This approach brings several advantages like easy scalability, and the possibility to address cumulative exposures to multiple stressors.

To support this approach, a case study was conducted to evaluate the long-term population exposure to PM2.5, NO2, and O3 in two European regions—Lombardy, Italy, and Germany—covering the period from 2013 to 2022 using the Copernicus Atmosphere Monitoring Service (CAMS) European Air Quality Reanalysis data. These datasets provide consistent and reliable estimates of pollutant concentrations, enabling a detailed evaluation of exposure by considering a static (residential) and dynamic (commuting habits included) population. The analysis integrated commuting data from national statistical institutes as well as the remote-sensing-derived global settlement mask, the World Settlement Footprint. The results highlight the significant disparities between exposure while considering a static and a dynamic population, emphasizing the importance of accounting for mobility in health risk assessments. Furthermore, the study demonstrates widespread exceedances of the World Health Organization’s updated air quality guidelines, particularly for PM2.5, and underscores the spatial variability in exposure levels.

To further investigate these variations, the study proposes the use of spatial analysis techniques, particularly the Local Indicators of Spatial Association (LISA) to study the temporal evolution of air pollution hotspots and cold spots. By applying this method, the research aims to identify spatial clusters of pollutants such as particulate matter (PM2.5), nitrogen dioxide (NO2), and ozone (O3) as well as to produce multi-hazard maps of areas where hotspots of multiple pollutants converge. These analyses provide critical insights into regions with heightened health risks and inform strategies for mitigating exposure.

As an exploratory approach, the LISA spatial analysis is extended to additional areas worldwide, exploring NO2 column data from satellite missions such as the Ozone Monitoring Instrument (OMI) and Sentinel-5P TROPOspheric Monitoring Instrument (TROPOMI) . These provide consistent, worldwide air quality data, offering opportunities to overcome the limitations of traditional ground-based monitoring networks.