气候变化领域集成服务门户

Two separate air quality model methodologies for estimating background ozone levels over the western U.S. are compared in this analysis. The first approach is a direct sensitivity modeling approach that considers the ozone levels that would remain after certain emissions are entirely removed (i.e., zero-out modeling). The second approach is based on an instrumented air quality model which tracks the formation of ozone within the simulation and assigns the source of that ozone to pre-identified categories (i.e., source apportionment modeling). This analysis focuses on a definition of background referred to as U.S. background (USB) which is designed to represent the influence of all sources other than U.S. anthropogenic emissions. Two separate modeling simulations were completed for an April October 2007 period, both focused on isolating the influence of sources other than domestic manmade emissions. The zero-out modeling was conducted with the Community Multiscale Air Quality (CMAQ) model and the source apportionment modeling was completed with the Comprehensive Air Quality Model with Extensions (CAMx). Our analysis shows that the zero-out and source apportionment techniques provide relatively similar estimates of the magnitude of seasonal mean daily 8-h maximum U.S. background ozone at locations in the western U.S. when base case model ozone biases are considered. The largest differences between the two sets of USB estimates occur in urban areas where interactions with local NOx emissions can be important, especially when ozone levels are relatively low. Both methodologies conclude that seasonal mean daily 8-h maximum U.S. background ozone levels can be as high as 40 -45 ppb over rural portions of the western U.S. Background fractions tend to decrease as modeled total ozone concentrations increase, with typical fractions of 75-100 percent on the lowest ozone days (<25 ppb) and typical fractions between 30 and 50% on days with ozone above 75 ppb. The finding that estimates of background ozone are not strongly dependent on the technique applied lends credibility to this and earlier modeling work. Published by Elsevier Ltd.