About Margaret
Margaret Lawrimore joined U-Spatial in April 2026 as a Geospatial Developer, bringing a Ph.D. in Geospatial Analytics from North Carolina State University.
Her research applies big-data analytics, high-performance computing, and geostatistical methods to climate change adaptation, sustainable development, and environmental justice, frequently in direct partnership with local communities and government agencies.
Before joining U-Spatial, she was a Graduate Research Assistant at NCSU's Urban Systems Lab and a Graduate Research Intern with the Human Geography Group at Oak Ridge National Laboratory.
Key Projects and Initiatives
Creating Spatially Complete Zoning Maps Using Machine Learning
Margaret and her co-authors built an open-source machine learning framework, using a Hierarchical Random Forest algorithm, to predict zoning classifications for places where that data doesn't exist, then used it to produce North Carolina's first statewide, comprehensive zoning map. The model correctly predicted zoning within a county roughly 99% of the time. The project was published in Computers, Environment and Urban Systems.
The Safe Development Paradox of the United States Regulatory Floodplain
This study examined whether the official 100-year floodplain boundaries that determine U.S. flood insurance requirements and development restrictions actually reduce risk. The team's national analysis found the opposite: floodplain regulation can paradoxically drive more development into and near flood-prone areas.
Smart Zoning for Coastal Flood Adaptation and Resilience
Funded by North Carolina Sea Grant, this project partnered with the Town of Leland's Community Development Planning Department to assess how the town's zoning regulations intersected with current and future flood exposure, identifying specific areas where zoning changes could reduce coastal flood risk. Margaret co-authored the resulting summary report, submitted directly to the town.
Areas of Expertise
- Developing spatial machine learning algorithms and predictive models to analyze urban growth, zoning patterns, and landscape changes.
- Evaluating coastal and regulatory floodplains to model population mobility, future development trends, and climate resilience strategies.
- Leveraging high-performance computing (HPC) clusters, cloud platforms like Google Earth Engine, and advanced Python/R libraries to process and analyze spatial data.
- Translating complex geostatistical research into visual tools, community reports, and actionable insights for urban planners and government partners.
Work with Margaret
Reach out, consultations are free for the U of M community.