Seongwon Im, Ph.D.

I am an applied statistician working on public-interest problems at the intersection of environmental science, public health, and decision-making. My work centers on population-level questions where uncertainty is inherent and statistical thinking can meaningfully inform real-world choices. I focus on Bayesian statistics, particularly Bayesian nonparametric modeling, statistical computing, and time-varying exposure analysis, with an emphasis on flexible yet interpretable methods for complex, high-dimensional data common in environmental and health research. A core aim of my research is to develop accessible statistical methods and software that support interdisciplinary collaboration, reproducible science, and evidence-informed decisions.

• Ph.D., Colorado State University, Fort Collins, CO
• M.S., Colorado State University, Fort Collins, CO
• B.S., St. Lawrence University, Canton, NY

View CV

Research interests

• Bayesian methods & Nonparametric modeling
• Distributed lag models & Time-varying exposure analysis
• High-dimensional & Heterogeneous mixture models
• Applications in environmental health & Public health

Publications

• Im, S., Wilson, A., & Mork, D. (2025). Structured Bayesian Regression Tree Models for Estimating Distributed Lag Effects: The R Package dlmtree. The R Journal, 17(1), 136–159. doi:10.32614/RJ-2025-007
• Weller ZD, Im S, Palacios V, Stuchiner E, von Fischer JC. 2022. Environmental Injustices of Leaks from Urban Natural Gas Distribution Systems: Patterns among and within 13 U.S. Metro Areas. Environmental Science & Technology, 56(12):8599–8609. doi:10.1021/acs.est.2c00097

R package: dlmtree

• Bayesian treed distributed lag models for time-varying exposure-response relationships
• Incorporates mixture exposures, heterogeneity, non-linearity, and multiple response types.

Bayesian modeling MCMC Rcpp Package development HPC

Urban Natural Gas Leaks and Environmental Injustice

• Spatial Bayesian models to estimate leak densities and environmental disparities across 13 U.S. metro areas.
• Integrates methane sensor data, GIS boundaries, and standardized census predictors to perform spatial analyses.

Bayesian inference Spatial modeling RStan Mapping Consulting

Invited & Oral Presentations

Posters & Contributed Talks