Month: May 2026

Geography According to Foundation Models

Editors: Krzysztof Janowicz, Rui Zhu, Gengchen Mai, Song Gao, Yingjie Hu, Zhangyu Wang, Ling Cai, Lauren Bennett

ISBN: 978-1-64368-659-2 (online), Open Access: https://ebooks.iospress.nl/volume/geography-according-to-foundation-models

As generative AI continues to evolve, the geographic community has an important role to play in shaping how these technologies represent space, understand place, and influence how people engage with the world. This edited volume, Geography According to Foundation Models, explores the opportunities and challenges faced as generative AI intersects with geographic knowledge, spatial reasoning, and Geographic Information Science (GIS); a discipline long concerned with the representation, analysis, and interpretation of geographic data to help us to understand the world around us. AI, which can generate, transform, manipulate, synthesize, and interpret various data modalities, has given rise to new opportunities for geographic knowledge discovery, while at the same time presenting challenges around trust, bias, brittleness, explainability, and pedagogy. The 14 chapters included here bring together the work of researchers studying how systems increasingly facilitated by AI are representing geographical processes. Contributions are grouped into 5 parts: foundations and frameworks; methods and applications; interpretability and evaluation; impacts on education; and trust, bias, and beyond, and cover topics which explore advancing the technical foundation of GeoAI and geospatial foundation models; leveraging large language models (LLM) for geographic information retrieval and behavior simulation; investigating how foundation models extract, perceive, and represent geographic information; the disruptive pedagogical implications of AI for geography and GIScience; and the socio-technical foundations of developing and adopting foundation models in geographic contexts. Exploring the intersection of AI with geographic knowledge, this volume will be of interest to geographers, data scientists, AI researchers, educators, and policymakers.

Sustainable human–wildlife coexistence requires a mechanistic understanding of the many ways that humans affect animals. However, progress is hampered by the lack of accessible data measuring the dynamic presence of people. Here, we leverage large-scale mobile-device data to disentangle how human mobility or presence and landscape modification differentially influence the use of geographic and environmental space for 37 mammal and bird species across the United States. Human presence affected more than 65% of species, with substantial variation across species. For ~60% of species that responded to human activities, the effects were interdependent—animals tended to react more strongly to human mobility/presence in less modified habitats. Our results demonstrate that human mobility/presence and landscape modification have complex combined effects on wildlife, which need to be considered for effective management.

Ruth Y. Oliver, Scott W. Yanco, Diego Ellis-Soto, Brett R. Jesmer, Juliet Cohen, Song Gao  et al. (2026) Interacting effects of human presence and landscape modification on birds and mammals. Science, 392,879-884.

All data, including environmental annotations, area-normalized human mobility data, and estimated area and niche size as well as code to reproduce results are available on the Open Science Framework: https://doi.org/https://doi.org/10.17605/OSF.IO/3UA2C

Code: https://github.com/GeoDS/Science-HumanMobilityAnimalEcology

Understanding how urban systems and traffic dynamics co-evolve is crucial for advancing sustainable and resilient cities. However, their bidirectional causal relationships remain underexplored due to challenges of simultaneously inferring spatial heterogeneity, temporal variation, and feedback mechanisms. Here we present a spatio-temporal causality framework that bridges correlation and causation by integrating spatio-temporal weighted regression with spatio-temporal convergent cross-mapping. Characterizing cities through urban structure, form, and function, the framework uncovers bidirectional causal patterns between urban systems and traffic dynamics across 30 cities on six continents. Our findings reveal asymmetric bidirectional causality, with urban systems exerting stronger influences on traffic dynamics than the reverse in most cities. Urban form and function shape mobility more profoundly than structure, even though structure often exhibits higher correlations. This does not preclude the reversed causal direction, whereby long-established mobility patterns can also reshape the built environment over time. Finally, we identify three causal archetypes: tightly coupled, pattern-heterogeneous, and workday-attenuated, which support city-to-city learning and inform context-sensitive strategies in sustainable urban and transport planning.

Zhang, Y., Hong, Y., Gao, S., & Raubal, M. (2026). Bidirectional yet asymmetric causality between urban systems and traffic dynamics in 30 cities worldwide. Nature Communications. https://www.nature.com/articles/s41467-026-71377-0

The code and data are publicly available at FigShare: https://doi.org/10.6084/m9.figshare.28656800