A Statistical Framework to Identify the Influence of Large-Scale Weather Events on Regional Air Pollution

Abstract

egional air pollution episodes occur as a result of increased emissions and prevalence of conducive meteorological conditions. The frequency of occurrence of such favorable conditions on a regional scale may be influenced by large-scale climatic events like ENSO and the Pacific decadal oscillation (PDO). The scarcity of measurements of criteria pollutants, especially ozone and particulate matter (PM), prior to the last 10?15-yr period, limits the scope of observing the influence of climate variability during recent decades on regional pollution levels. The authors propose a novel statistical framework to utilize available measurements and characterize synoptic influences on regional PM pollution in California?s Central Valley during 1998?2008. The identified target conditions are used to develop a classification scheme to scan historical climate datasets dating back to 1948. The procedure identifies exceedance-conducive days during 1950?98, when no PM2.5 measurements were available. Temporal patterns in seasonal frequency of these identified exceedance-conducive days are investigated for temporal patterns driven by ENSO and PDO.