【摘要】 基于CiteSpace可视化工具，对2010年1月至2025年2月的中国知网(CNKI)及Web of Science(WOS)核心数据库中的山洪灾害领域文献进行计量分析.通过对1 173篇中文文献和1 021篇英文文献的分析，分别从发文量、国际合作情况、核心机构、作者网络、关键词共现和关键词突现等多个维度进行分析.研究表明：国内外研究规模存在显著差异，CNKI数据库年均发文量略微波动，WOS数据库从2017年起发文量持续上升，在2024年达到峰值169篇，彰显国际关注度的提升.机构分析表明，中国水利水电科学研究院、水利部防洪抗旱减灾工程技术研究中心等机构在国内外均占据主导地位，国际协作以国内高校为主.关键词分析发现，国内研究集中于“临界雨量”“监测预警”等灾害防治与应用技术，国际研究则侧重于“空间预测”“气候变化”“机器学习”等模型创新及多源数据的交叉融合.关键词突显了国内热点从“山洪成因”转向“风险评价”,国际关注“机器学习”等技术手段.当前国际合作不足、多因子耦合数值模拟能力有限、跨学科协同合作待加强等问题，未来应着眼于构建全球数据共享网络，深化多学科协同的人工智能合作分析，推动山洪灾害领域研究迈向可持续发展道路.更多还原

【Abstract】 Based on the CiteSpace visualization tool, a bibliometric analysis was conducted on literature in the field of flash flood disasters from the China National Knowledge Infrastructure(CNKI) and Web of Science(WOS) core databases between 2010 and 2025. Analyzing 1,173 Chinese and 1,021 English publications, this study examines multiple dimensions including publication volume, international collaboration, core institutions, author networks, keyword co-occurrence, and keyword bursts. The research indicates significant differences in the scale of research between China and the international community. The annual publication volume in the CNKI database shows slight fluctuations, while the WOS database exhibits a continuous increase since 2017, peaking at 169 publications in 2024, highlighting the growing international attention. Institutional analysis reveals that the China Institute of Water Resources and Hydropower Research and the Engineering Technology Research Center of Flood and Drought Prevention and Mitigation, Ministry of Water Resources, among others, play dominant roles both domestically and internationally. International collaboration primarily involves domestic universities. Keyword analysis finds that domestic research focuses on disaster prevention and application technologies such as “critical rainfall” and “monitoring and early warning, ” while international research emphasizes model innovation and the integration of multi-source data, such as “spatial prediction, ” “climate change, ” and “machine learning.” Keyword burst detection reveals a shift in domestic hotspots from “flash flood causes” to “risk assessment”, and international focus on technological means like “machine learning.” Current challenges include insufficient international collaboration, limited capabilities in multi-factor coupled numerical simulation, and a need for enhanced interdisciplinary cooperation. Future efforts should focus on building a global data-sharing network, deepening interdisciplinary collaboration and AI-assisted analysis, and promoting sustainable development in flash flood disaster research.更多还原