A multi-model assessment for the 2006 and 2010 simulations under the Air Quality Model Evaluation International Initiative (AQMEII) phase 2 over North America: Part I. Indicators of the sensitivity of O3 and PM2.5 formation regimes

Abstract Under the Air Quality Model Evaluation International Initiative, Phase 2 (AQMEII-2), three online-coupled air quality model simulations, with six different configurations, are analyzed for their performance, inter-model agreement, and responses to emission and meteorological changes between 2006 and 2010. In this Part I paper, we focus on evaluating O 3 and PM 2.5 indicator-based analyses, which are important in the development of applicable control strategies of O 3 and PM 2.5 pollution in different regions worldwide. The O 3 indicators agree on widespread NO x -limited and localized VOC-limited conditions in the U.S. The NO y and O 3 /NO y indicators overpredict the extent of the VOC-limited chemistry in southeast U.S., but are more robust than the H 2 O 2 /HNO 3 , HCHO/NO y , and HCHO/NO 2 indicators at the surface, which exhibit relatively more inter-model variability. The column HCHO/NO 2 indicator is underpredicted in the O 3 and non-O 3 seasons, but there is regional variability. For surface PM 2.5 indicators, there is good inter-model agreement for the degree of sulfate neutralization; however there are systematic underpredictions in the southeast U.S. There is relatively poor inter-model agreement for the less robust adjusted gas ratio indicator, which is largely overpredicted in the summer and both under- and overpredicted in winter in the southeast U.S. There is good inter-model agreement for the O 3 indicator sensitivities, indicating a predominant shift to more NO x -limited conditions in 2010 relative to 2006. There is less agreement for PM 2.5 indicator sensitivities, which are less robust, while indicating shifts to either regime due to different responses of aerosol treatments to changes in emissions and meteorology.