Geospatial data science and wildfire modeling at the University of Montana.
I'm a graduate research assistant at the University of Montana, working toward a Master of Science in Computer Science (expected August 2028). My research combines machine learning with wildland fire behavior and spread modeling, funded through the National Science Foundation's SMART FIRES project.
I come from a forestry background, with wildland fire experience, most recently as a forester for the Confederated Salish and Kootenai Tribes. That is why I care about geospatial data. Nearly every fire question is a spatial one: where it starts, what it burns through, and how terrain and weather push it. Maps, satellite imagery, and models let me ask those questions across a whole landscape.
Today I focus on testing how well wildfire models match real fires. My QUIC-Fire poster, presented at the SMART FIRES all-hands meeting in September 2026, used statistical analysis to validate the model against real wildfire observations.
Currently: I prepare model inputs from field and weather observations, run simulations, and turn the output into maps and figures that answer a question. That includes QUIC-Fire input pipelines, resolution and thread benchmarking, and the mapping tools around model runs. I'm also looking at where natural wildfire ignitions concentrate by fuel type across ecoregions of the contiguous United States, using 30+ years of federal fire occurrence records.
Programming and data: Python (GeoPandas, rasterio, xarray), Java, C, R, and SQL
Geospatial: Google Earth Engine, ArcGIS Pro, QGIS, CloudCompare, remote sensing, and point cloud processing.
Fire behavior: QUIC-Fire, FlamMap, and FARSITE
I'm interested in connecting with fire agencies, research labs, and companies working on wildfire risk and modeling.