The U.S. Clean Air Act identifies six criteria air pollutants considered harmful to public health and the environment. One of them is fine particulate matter (PM2.5), a mixture of particles and droplets in the atmosphere with a diameter of 2.5 microns or less. Each year the U. S. EPA examines fine PM trends and sources, and it compiles its concentration and composition measurements from monitoring stations located across the United States that provide the speciation of fine PM. Fine PM has been related to adverse health effects in animals and humans, including respiratory and cardiovascular disease morbidity and mortality
The aim of this work is to propose a nationwide spatio-temporal model for prediction and modelling of speciated fine particulate matter to have a better understanding of the composition and sources of PM2.5. For that, we develop and implement a spatial temporal factor analysis that provides a very broad, flexible multivariate approach to studying the spatio-temporal patterns of PM2.5 mass and its components, while characterizing its potential sources. The results from this work could have a significant impact in air quality policy and management for fine PM, in particular by understanding the space-dependent composition and sources of fine PM.
Keywords: Factor analysis; Space-time model; Particulate matter
Biography: Dr. Esther Salazar obtained her PhD degree from Federal University of Rio de Janeiro, Brazil. She worked as a Postdoctoral Researcher at Statistical and Applied Mathematical Sciences Institute (SAMSI), USA and, currently, she works as a Postdoctoral Associate at Duke University, USA. Her areas of expertise include spatio-temporal modeling, dimension reduction problems and computational statistics.