Mapping global trends in landslide susceptibility research using geospatial technologies

Abstract

This study maps the global landscape of geospatial-based landslide susceptibility research through a bibliometric analysis of Scopus-indexed publications from 2015 to 2025. OpenRefine was used for metadata cleaning and VOSviewer version 1.6.20 was applied for keyword co-occurrence and country co-authorship mapping. A total of 2,002 articles were identified, with a clear increase in publication output after 2019. The analysis identified five main thematic clusters: remote sensing and geospatial technologies, machine learning-based vulnerability modeling, geotechnics and slope dynamics, multi-hazard approaches and climate adaptation, and GIS-based empirical approaches. China, India, Italy, and the United States emerged as the main contributors in publication output, citation performance, and collaboration strength, while Remote Sensing, Landslides, and Natural Hazards were the leading publication outlets. These findings suggest an emerging shift from conventional GIS-based mapping to an integrated Geospatial-AI approach, supported by big earth data, cloud computing, and artificial intelligence. This study highlights the need for cross-regional collaboration, multi-source data integration, and transparent predictive models for environmental risk reduction.