Comparison of remote sensing and fixed-site monitoring approaches for examining air pollution and health in a national study population

作者：Prud' homme Genevieve; Dobbin Nina A; Sun Liu; Burnett Richard T; Martin Randall V; Davidson Andrew; Cakmak Sabit; Villeneuve Paul J; Lamsal Lok N; van Donkelaar Aaron; Peters Paul A; Johnson Markey^*

来源：Atmospheric Environment, 2013, 80: 161-171.

DOI：10.1016/j.atmosenv.2013.07.020

• Air pollution
• Satellite remote sensing
• Regulatory monitoring
• Asthma
• Allergy
• Population health
• Epidemiology

Satellite remote sensing (RS) has emerged as a cutting edge approach for estimating ground level ambient air pollution. Previous studies have reported a high correlation between ground level PM2.5 and NO2 estimated by RS and measurements collected at regulatory monitoring sites. The current study examined associations between air pollution and adverse respiratory and allergic health outcomes using multi-year averages of NO2 and PM2.5 from RS and from regulatory monitoring. %26lt;br%26gt;RS estimates were derived using satellite measurements from OMI, MODIS, and MISR instruments. Regulatory monitoring data were obtained from Canada%26apos;s National Air Pollution Surveillance Network. Self-reported prevalence of doctor-diagnosed asthma, current asthma, allergies, and chronic bronchitis were obtained from the Canadian Community Health Survey (a national sample of individuals 12 years of age and older). Multi-year ambient pollutant averages were assigned to each study participant based on their six digit postal code at the time of health survey, and were used as a marker for long-term exposure to air pollution. %26lt;br%26gt;RS derived estimates of NO2 and PM2.5 were associated with 6-10% increases in respiratory and allergic health outcomes per interquartile range (3.97 mu g m(-3) for PM2.5 and 1.03 ppb for NO2) among adults (aged 20-64) in the national study population. Risk estimates for air pollution and respiratory/allergic health outcomes based on RS were similar to risk estimates based on regulatory monitoring for areas where regulatory monitoring data were available (within 40 km of a regulatory monitoring station). RS derived estimates of air pollution were also associated with adverse health outcomes among participants residing outside the catchment area of the regulatory monitoring network (p %26lt; 0.05). %26lt;br%26gt;The consistency between risk estimates based on RS and regulatory monitoring as well as the associations between air pollution and health among participants living outside the catchment area for regulatory monitoring suggest that RS can provide useful estimates of long-term ambient air pollution in epidemiologic studies. This is particularly important in rural communities and other areas where monitoring and modeled air pollution data are limited or unavailable.

• 出版日期2013-12