The Landscape of LLM-Powered Geoanalysis

The emergence of large language models (LLMs) has opened up new possibilities for geoanalysis, spanning geospatial and geospatio-temporal data analysis, making it easily accessible to non-GIS experts via their natural language interface. In this study, we present the first systematic literature review (SLR) of LLM-powered geoanalysis tools.We analyzed 17 peer-reviewed studies published between 2022 and 2025. Our review synthesizes current developments along several key dimensions. Based on our review, we propose a taxonomy of geoanalysis data categories, classify the types of geoanalysis supported, and identify application domains where these techniques are being applied. Additionally, we examine the LLM models, programming languages, and libraries employed, as well as the extent to which systems offer open data and source code access.We further analyze evaluation strategies by comparing performance metrics and methodological approaches across studies. Finally, we discuss common challenges, such as LLM hallucination, tool chaining limitations, and data compatibility issues, in the context of geoanalysis and summarize the solutions proposed in the literature. This paper provides a valuable and comprehensive overview of the landscape of LLM-powered geoanalysis tools and frameworks to aid our research and development community in adopting LLMs for geoanalysis.