High Spatiotemporal Resolution PM2.5 Speciation Exposure Modeling in California
Contact
Principal Investigator/Author: Jun Wu
Contractor: University of California, Irvine
Sub-contractor: Meredith Franklin, University of Southern California
Contract Number: 21RD006
Project Status: Active
Relevant CARB Programs: Exposure, California State Implementation Plans
Topic Areas: Air Pollution Exposure, State Implementation Plans (SIPs), Particulate Matter (PM), PM2.5
Research Summary:
PM2.5 is a mixture of many different chemicals. While many air quality monitors measure the total mass of PM2.5 in the air, there are relatively few measurements of the chemical makeup of PM2.5, referred to as PM2.5 chemical speciation. Understanding PM2.5 chemical speciation is important for determining the most important sources of PM2.5 in our air, as well as understanding some of the health effects of PM2.5. The lack of existing PM2.5 chemical speciation data makes this sort of analysis challenging, if not impossible.
The primary objective of this research contract is to develop methods for estimating PM2.5 chemical speciation across the State. Advanced machine learning models will be created using a variety of data as inputs, including ground-based monitors, meteorological data, satellite data, air quality models, and other land use variables. The researchers will use the outputs from their machine learning model to characterize human exposure levels across various sociodemographic factors, such as age, gender, race/ethnicity, and household income, and will recommend pathways to refine existing PM2.5 mitigation strategies.
Keywords: PM2.5; PM2.5 chemical speciation; PM2.5 composition; satellites; machine learning