Data Integration Model for Air Quality: A Hierarchical Approach to the Global Estimation of Exposures to Ambient Air Pollution

Abstract

Summary. Air pollution is a major risk factor for global health, with 3 million deaths annually being attributed to fine particulate matter ambient pollution (PM 2:5 ). The primary source of information for estimating population exposures to air pollution has been measurements from ground monitoring networks but, although coverage is increasing, regions remain in which monitoring is limited. The data integration model for air quality supplements ground monitoring data with information from other sources, such as satellite retrievals of aerosol optical depth and chemical transport models. Set within a Bayesian hierarchical modelling framework, the model allows spatially varying relationships between ground measurements and other factors that estimate air quality. The model is used to estimate exposures, together with associated measures of uncertainty, on a high resolution grid covering the entire world from which it is estimated that 92% of the world's population reside in areas exceeding the World Health Organization's air quality guidelines.

Stringent Emission Control Policies Can Provide Large Improvements in Air Quality and Public Health in India

“…The model underestimated ambient PM 2.5 concentrations near the Thar Desert and in the central IGP, which was also found in previous simulations over India (Kumar et al, 2014). Ambient PM 2.5 concentrations were similar to those from GBD2015 (Shaddick et al, 2018), apart from our model simulating lower estimates in the central and western IGP (Conibear et al, 2018). Further model evaluation for aerosol optical depth against the aerosol robotic network surface measurements showed similar close agreement (normalized mean bias = 0.09; Conibear et al, 2018).…”

“…This premature mortality estimate total is 9% lower than reported in Conibear et al (2018) primarily due to the slightly lower risk estimates for cardiovascular diseases from the GBD2016 exposure-response function relative to the GBD2015. Our mortality estimate is 13% lower than the estimate from the GBD2016 (GBD 2016 Risk Factors Collaborators, 2017), where the difference results from a combination of slightly lower population density from IFs over India, lower baseline mortality rates at higher ages for cardiovascular diseases, and our slightly lower PM 2.5 concentrations over the central and western IGP (Conibear et al, 2018;Shaddick et al, 2018).…”

Quantifying the Contribution to Uncertainty in Mortality Attributed to Household, Ambient, and Joint Exposure to PM2.5From Residential Solid Fuel Use

“…The integrated exposure‐response function published in Cohen et al () includes 1,000 sets of coefficients fit to the health data. PM Amb : Ambient PM 2.5 exposure concentration from all sources (available from https://www.stateofglobalair.org/) presented as an annual, population‐weighted mean at the country level. Mean PM 2.5 and 95% confidence levels (also country level) about the mean are included in Cohen et al () and based on methods described in Shaddick et al (). Following the assumptions in Cohen et al (), the ambient PM 2.5 concentrations are assumed to represent exposure. %PM A‐SFU : Percent of the PM Amb originating from SFU emissions.…”

Visualization in Bayesian Workflow

“…In order to better discuss the ways visualization can aid a statistical workflow we consider a particular problem, the estimation of human exposure to air pollution from particulate matter measuring less than 2.5 microns in diameter (PM 2.5 ). Exposure to PM 2.5 is linked to a number of poor health outcomes, and a recent report estimated that PM 2.5 is responsible for three million deaths worldwide each year (Shaddick et al, 2017).…”