Preferentially Monitored Air Pollution Fields: The Effect on Estimates of Annual Averages and Relative Health Risks

• Activity Number: 222
• Type: Invited
• Date/Time: Monday, August 10, 2015 : 2:00 PM to 3:50 PM
• Sponsor: ENAR
• Author(s): James Zidek* and Gavin Shaddick
• Companies: The University of British Columbia and University of Bath
• Keywords: air pollution ; high dimensional data ; INLA ; health effects ; preferential sampling ; spatio--temporal epidemiology

Abstract

In the 1960s, around 2000 sites in the UK monitored black smoke (BS) air pollution due to concerns about its effect on public health that were clearly demonstrated by the famous London fog of 1952. Abatement measures led to a decline in the levels of BS and hence a reduction in the number of monitoring sites to less than 200 by 1996. The authors' published case study shows that sites were removed preferentially, leading to exaggerated estimates of pollution levels. This talk will describe methods for mitigating the effects of that overestimation. In particular, it will show in an empirical study, that annual BS concentration averages may be substantially overestimated while paradoxically, the relative risk of environmental health outcomes may be underestimated along with appropriate adjustments. As a secondary feature, the talk will describe how the investigators handled the large p and small n problem they encountered due to the large number of monitoring sites in the BS network - standard geostatistical and Bayesian approaches do not work. The research has important general implications for setting regulatory standards and for designing monitoring networks