Fostering Air Pollution Health Risk Assessment in Urban Areas by means of the World Settlement Footprint

In 2020, 56% of the global population lived in urban areas and in Europe this figure is as high as 75%, with a steadily increasing trend (The World Bank). According to the European Environmental Agency’s (EEA) Air Quality Report, 20,600 premature deaths in Europe in 2018 were attributable to air pollution, with the highest proportion of people exposed to levels above the World Health Organization’s Air Quality Guidelines (WHO_AQG) living in cities (European Environmental Agency 2020). This is just one of multiple, mutually interacting phenomena constituting the complex and constantly developing urban ecosystem. To develop a better understanding of the related processes and interactions, enhanced tools are needed to enable planners and decision makers to effectively reduce the exposure of the urban population to air pollution. The causal relationship between air pollution and the exacerbation of health outcomes, ranging from acute respiratory and cardiovascular diseases to chronic illnesses, has been comprehensively documented (Beelen et al. 2014, Di et al. 2017, Liu et al. 2021). Subsequently, a quantification of the vulnerability to health impairments of different target groups in the form of concentration-response functions has been provided by key health authorities, most recently by the WHO-AQG, published in 2021. Vulnerability alone is not sufficient for a comprehensive quantification of risk. In fact, an accurate risk assessment must include three essential components: the assessment of the hazard (i.e., the air pollution concentration), the vulnerability of the individual (i.e., the dose-response relationship), and the probability of exposure (United Nation Office for Disaster Risk Reduction 2017). This task can be easily accomplished at the individual level using wearables, microsensors, or mobile data (Zou et al. 2009, Dewulf et al. 2016). Despite the valuable results of studies implementing these systems, they can only provide a temporally and spatially limited snapshot of the complex reality. Hence, the challenge is to fill this data and monitoring gap. To do so, a method is necessary, that allows for an exposure estimation at local scale and, consequently, a quantification of the health risk from air pollution that is reliable, easy scalable and applicable to different areas of interest, possibly worldwide. In our study, we propose a top-down approach, exploiting the wealth of environmental data from remote sensing, in-situ measurements and modelling to assess the increase in health risk due to air pollution exposure in the urban environment. The Copernicus Atmospheric Monitoring Service (CAMS) multi-year reanalysis data (Marécal et al. 2015) is used to assess the urban air quality. The exposure estimation is obtained by exploiting the recent World Settlement Footprint (WSF) 2019, a global settlement extent mask derived at 10m spatial resolution by jointly exploiting multitemporal Sentinel-1 and Sentinel-2 imagery (Marconcini et al. 2021). The percent settlement area derived at 100m resolution from the WSF2019 is used as a proxy for the probability of human presence in the different city compartments during specific time frames. The results are risk maps, revealing the increase in health risk due to air pollution in the different city compartments. Quantifying the total load is also a fundamental metric for policy makers. To this purpose, the novel WSF2019 population layer is employed, which estimates for each 10m resolution pixel marked as settlement in the WSF2019 the corresponding number of inhabitants. Specifically, this is obtained by proportionally redistributing population figures available at the finest possible administrative level by means of the local imperviousness (i.e., a reliable proxy for the built-up density), as well as, land-use information gathered from OpenStreetMap. In this way, an additional metric – the health burden index – that considers the number of people affected by different levels of risk can be calculated. The approach presented here aims at providing policy makers with high quality information to support the sustainable urban development. The final goal is to increase the resilience of urban areas against the negative effects of air pollution. Quality information result from a trade-off between a large amount of input data and the accuracy of the metric provided, with the scope of making this method easily applicable to different geographical contexts. Our work can potentially support the research on the causality relationships between air pollution and health outcomes by exploiting an ecological design. Future studies should focus on validating the results using collected health data