Remote sensing in river obstruction research: A bibliometric analysis integrated with large language model

River obstructions affect hydrological, water quality, and ecosystem, making effective monitoring and management strategies essential. Remote sensing has advanced obstruction research, yet traditional literature review methods face challenges due to the increasing volume of publications. To address this, this study integrated bibliometric analysis with large language models (LLMs) to provide a comprehensive review of the role of remote sensing in river obstruction research. Analyzing Web of Science and Scopus data revealed a notable rise in publications since 2010. LLM-based topic classification achieved high accuracy, with an average precision of 88.1 % and recall of 86.1 %. The analysis further identified hydrological impacts (39.05 %) and ecological impacts (26.63 %) of obstructions as the most studied. Study areas were identified in 44 % of publications, with the Yangtze (25.91 %) and Mekong (11.20 %) River Basins being prominent. Temporal analysis (2012–2025) revealed a research shift from dam operation to increasing emphasis on ecological degradation and water quality deterioration, particularly in rapidly developing basins like the Yangtze and Mekong. Remote sensing has proven instrumental in detecting river obstructions and analyzing hydrological and ecological impacts. This study proposes a novel framework that combines bibliometric analysis with LLMs, demonstrating the potential of LLMs in analyzing science literature. The framework uncovers basin-specific hot topics across global river basins and provides fresh perspectives on research trends and hotspots related to remote sensing technologies in river obstruction studies.