PhD candidate in Environmental Sciences & Engineering at UNC-Chapel Hill, working at the intersection of atmospheric science, geostatistics, and machine learning. I build geospatial data-fusion methods that turn monitors, satellites, and model output into high-resolution maps of air pollution for health and policy applications.
- Data fusion for air quality: combining ground monitoring, satellite observations, and chemical transport model output (CMAQ, CAMx) with Bayesian Maximum Entropy to produce high-resolution pollutant maps.
- Exposure and environmental justice: quantifying exposure disparities across demographic groups, including hazardous air pollutants in the US Gulf States.
- Global ozone: fusing TOAR-II observations with RAMP-corrected model output for global-scale ozone mapping.
- Machine learning for environmental science: stacked ensemble models for pollutant prediction, plus agentic LLM workflows for scientific code.
Python (xarray, NumPy, pandas, scikit-learn, PyTorch), MATLAB, R, SQL, Fortran, NetCDF and geospatial stacks, ArcGIS/QGIS, Git
| Project | What it does |
|---|---|
| RAMP-Py | Processing, RAMP correction, and ensemble modeling of global ozone data |
| gO3_II_mo | Fuses TOAR-II observations with RAMP-corrected CTM output for high-resolution global ozone estimates |
| bme-sbtex | BME mapping of SBTEX using monitoring data and CAMx, part of HAPI |
| BMEGUI | Maintaining and modernizing the BMEGUI spatiotemporal mapping toolkit |
Open to data science and research roles. Happy to talk about air quality modeling, spatiotemporal statistics, or scientific tooling.
Website: praful-dodda.github.io