Investigating particulate matter in air pollution across the United States, and formulating possible interventions to reduce its impact on public health
The data used in this research comes from the Atmospheric Composition Analysis Group. Ground-level fine particulate matter (PM2.5) total and compositional mass concentrations for North America, China, Europe, and other regions are publicly available at this link. We used the monthly data from the V4.NA.02 dataset, restricting the boundaries to the continental US.
As of October 2021, my assistance included performing preliminary data exploration by visualizing the spatial data in R and running clustering algorithms to determine the spatial and temporal relationships of the different elements composing the particulate matter (black carbon, dust, sulfate, etc.).
As of February 2022, I have started looking into causal inference using methods like synthetic control and Bayesian structural time series (BSTS), the latter of which is not a causal effects method but can be extended to causal applications. These methods will allow us to evaluate the impact of various pollution policies or other time-relevant events on the pollution mixture / concentration in an area.