A21A-0069
Using Positive Matrix Factorization to Investigate Sources of VOCs in Bakersfield

Tuesday, 15 December 2015
Poster Hall (Moscone South)
Shelby Eckel Scola1, Jason Schroeder2 and Donald Ray Blake2, (1)Providence College, Providence , RI, United States, (2)University of California Irvine, Irvine, CA, United States
Abstract:
Bakersfield, California consistently ranks among the nation's top three most polluted cities in terms of both ozone and particulate pollution. An important step in developing control strategies for the mitigation of ozone is determining the contribution of various emission sources of ozone precursors, such as volatile organic compounds (VOCs), in Bakersfield. During the SARP 2015 campaign, whole air samples were collected over Bakersfield and other Central Valley emission sources - including oil and natural gas sites and agricultural areas - and analyzed by gas chromatography. To estimate the influence of oil and natural gas sources on air in Bakersfield, the ratio of i-pentane to n-pentane was used. Use of this ratio demonstrated mixed urban and oil and natural gas influences in the Bakersfield samples. To better identify and quantify the contributions of oil and natural gas fields as well as other regional emission sources on Bakersfield air quality, positive matrix factorization (PMF) was applied to whole air samples taken within the planetary boundary layer in the Central Valley. PMF generated three interpretable factors: an urban source, an oil and natural gas source, and an agricultural/biogenically-influenced source. The contribution of each of the three sources on the mixing ratios of C1-C7 alkanes and isoprene was calculated. Of the non-methane VOCs in Bakersfield, it was determined that 45% originated from the oil and natural gas sites, 34% from the agriculture/biogenically-influenced sources, and 21% from urban areas. Furthermore, it was observed that there was a better agreement between PMF results and propane to ethyne ratios than with i-pentane to n-pentane ratios.