Using 2D Wavelets to Blend Grid- and Point-Based Air Quality Model Output with Ambient Monitor Observations: Resolving the Intra-Urban Air Pollution Field

• Activity Number: 611
• Type: Contributed
• Date/Time: Thursday, August 6, 2009 : 10:30 AM to 12:20 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): James L. Crooks*+ and Vlad Isakov
• Companies: U.S. Environmental Protection Agency and U.S. Environmental Protection Agency
• Keywords: Bayesian Statistics ; Wavelets ; Environmental Statistics ; Air Quality ; Computer Models

Abstract

Among environmental epidemiologists there is a desire to improve the accuracy of ambient air pollution exposure estimates by resolving the intra-urban concentration field. Unfortunately, monitor networks are sparse, costly to maintain, and are often designed to miss local features. Grid-based regional photochemical simulations, on the other hand, cannot resolve intra-urban structure, while plume dispersion models, which can resolve this structure, cannot track long-range transport or chemical reactions. We present a new Bayesian method that blends these three sources of information to resolve the intra-urban pollution field. This method represents the latent field using a 2D wavelet basis, which allows direct, efficient comparison to data at multiple levels of resolution. We use it to estimate daily Baltimore-area Benzene concentration at census block-group level for the year 2002.