Historical Population Exposure to Fine Particulate Matter Extracted by Spatiotemporal Interpolation from 2005 to 2015 across the Contiguous U.S.

摘要

Fine particulate matter (PM_(2.5)) isassociated with increased risk of mortality and respiratory diseases. Epidemiological studies consistently show an association between atmospheric particle pollution and the number of deaths from cancer, and cardiovascular and respiratory diseases. To further investigate the links between PM_(2.5) and adverse health effects, it is imperative to estimate PM_(2.5) exposure in a continuous space-time domain. To estimate PM_(2.5) density and distribution, spatial interpolation methods have been well developed to estimate values at unknown locations, but few approaches have concurrently considered the contribution of data in the time dimension. Integrating space and time simultaneously is anticipated to yield better interpolation results than treating them separately for typical GIS applications (Li et al., 2012). Unfortunately, there are far fewer models for spatiotemporal interpolation compared with spatial interpolation (Li et al., 2004, Li et al., 2016), especially in the application of air pollution where data varies within a large time domain. In addition, continuous exposure to a higher level of PM_(2.5) may have a much severer impact on public health than intermittent exposure. Hence, it is important to consider long-term, temporal variation of air pollution exposure across multiple years. In this study, we applied innovative spatiotemporal interpolation techniques based on cloud computing to process large amount of historical PM_(2.5) data. Results include monthly PM_(2.5) values at census block group level in the past 10 years from 2005 to 2015. We linked the pollution levels with the demographics across the U.S. The spatiotemporal exposure levels to population across the contiguous U.S. were quantified and visualized in our results. In summary, this study explored population exposure to PM_(2.5) in the contiguous U.S. from 2005 to 2015 and provided public health implications based on PM_(2.5) exposure at both geography and time dimensions.