St. James Ballroom
Investigating the Underlying Structure of Particulate Matter Concentrations (303809)
*Eduardo Montoya, California State University, Bakersfield
Keywords: Statistical Modelling, Functional Data, Air Pollution
The air pollutants that pose among the highest risk to California are ground level ozone and particulate matter (PM). However, reducing air pollution concentrations to acceptable levels remains an on-going challenge in California where eight of its cities rank in the top 10 of the highest levels of particulate matter pollution in the US. This challenge is becoming more complex as air quality is susceptible to climate change. The connection between PM and climate change is not well established. We propose to add to the literature by exploring the underlying data structure of the size variation and time variation of PM and its association with regimes of Pacific climate using functional data methodologies using R. Specifically, we apply curve alignment to separate the time and size variation of the PM, and functional principal component analysis is applied on the resulting variations to extract their dominant modes of variability. We then investigate how these dominant modes are associated with spatial location and certain patterns of atmospheric and oceanic variability over the Pacific Ocean.