Investigation of Chicago Air Pollution and Meteorological Influence on Pollution Development: Physical Measurements and Multivariate Statistical Analyses

The goal of this project was to evaluate air pollutants and meteorology in Chicago, Illinois during summers 2010, 2011, and 2012. Sixty-six collection days of air pollutant and meteorological data were completed at the Loyola University Chicago sampling station atop Mertz Hall. Trace gases, nitrogen oxides and ozone, as well as the water-soluble fraction of aerosol material were the air pollutants of focus in this study. Data analysis showed that chloride, nitrate, sulfate, and oxalate were the most frequently quantified water-soluble ions in aerosol samples, followed by acetate, formate, potassium, phosphate, calcium, and magnesium ions. Ionic ratios showed that stationary sources, vehicular emissions, and secondary production contributed to ions\u27 presence in the local atmosphere, however, each summer had a different combination of sources. Each summer had a different dominant wind direction profile, which contributed to the variability in ion and trace gas concentrations in the local atmosphere, as well as temperature, wind speed, and humidity variability. Collection days were classified based on wind direction to define non-lake breeze, lake breeze, and variable days. The onset of a lake breeze varied from 0900 to 1245 local time (LT) and onset time was significantly affected by both morning wind speed and temperature. The majority of air pollutants were not significantly affected (Student t test) by the onset of lake breeze circulation. However, ozone, sulfate, nitrate, and chloride concentrations were significantly different (p \u3c 0.05) on lake breeze days compared to non-lake breeze days. Summer 2010-2012 data was compared to summer 2002-2004 data collected at the same location. Nitrogen oxides and sulfate concentrations decreased, however, ozone and many water-soluble ions increased in concentration since the earlier study. Identifying point sources of air pollution and calculating backward air parcel trajectories proved valuable in understanding pollution variability. Additionally, canonical correlation analysis and principal component analysis were independently applied to sets of air pollutant and meteorological data from summers 2002-2004 and 2010-2012. Both methods uncovered multivariate relationships between pollutant-pollutant and pollutant-meteorological variables. With this information, a better understanding of the complex nature of pollutants and their dependence on meteorology was achieved