Research
The Van Berkel Lab develops geospatial theory, methods, and decision-support approaches to understand how communities and landscapes change under environmental and societal pressure. Across projects, we connect spatial analysis with substantive questions about climate adaptation, migration, urban change, conservation, and environmental governance.
Climate adaptation, migration & urban futures
How will climate-related migration reshape cities, land use, infrastructure, environmental quality, and inequality?
We combine land-change modeling, demographic scenarios, climate information, and stakeholder engagement to examine how population movement may alter urban systems. Current work spans the North American Great Lakes and the Lake Victoria Basin, with particular attention to housing, infrastructure, environmental exposure, and adaptation planning.
Geospatial artificial intelligence
How can AI extract meaningful and responsible information from large spatial and visual datasets?
Our work develops and evaluates computer vision, vision-language models, remote sensing, and spatial machine learning for environmental and urban applications. Current research includes residential property conditions and blight in Detroit, landscape perception, wildlife and conservation communication, and large-scale ecological inference.
Participatory GIS & decision support
How can local knowledge and community priorities become part of spatial analysis rather than remain separate from it?
We develop participatory mapping systems that allow residents and practitioners to identify risks, preferences, priorities, and tradeoffs. PIVOT and related platforms connect these inputs with spatial models, enabling participatory data to directly inform scenario analysis and planning.
Flood resilience & actionable climate knowledge
How can climate, hydrologic, and community knowledge be integrated to support more equitable flood planning?
Through GLISA and CHIRRP/PUMP-COR, we study the production and use of actionable climate knowledge in Great Lakes communities. This work links climate science, hydrologic modeling, participatory GIS, and co-production in places including Benton Harbor, Milwaukee, Detroit, and other regional cities.
Landscape perception, conservation & human–environment relationships
How do people perceive, value, and interact with landscapes, and how can those relationships inform conservation and planning?
We use social media, surveys, landscape visualization, spatial analysis, and ecological data to study cultural ecosystem services, recreation, urban forests, wildlife engagement, and environmental perception. This research connects human experience with landscape and conservation outcomes.
Methods
Our work commonly draws on GIS · spatial statistics · R · Python · remote sensing · machine learning · computer vision · large language models · agent-based and cellular-automata modeling · social media analytics · street-level imagery · participatory GIS · interactive visualization · Shiny/Leaflet · qualitative and mixed methods.