【摘要】 本文以2024年9月17日发生在华东四省的龙卷爆发事件为例，发展并实践了一套融合多源信息与灾情特征的系统性调查方法。该方法涵盖舆情收集与分析、调查路径规划、现场调查技巧、无人机航拍技术以及灾情的分析评估等关键环节。应用此方法，灾情调查组在三天内高效完成了10处舆情点的实地验证，确认其中9处由11个独立的龙卷所致(包括EF1级8个、EF0级3个),并成功识别出一处虚假舆情。本文系统阐述了基于条状痕迹、辐合倒伏与灾情狭道等关键地面特征判别龙卷移向的技术，并据此制定了标准化的无人机航拍流程，有效克服了弱龙卷地面痕迹模糊、雷达探测局限等调查难点。通过综合龙卷漏斗视频、地面灾情特征与雷达速度对等信息，研究完成了对所有龙卷的精确判识与定损评估，构建了包含路径、强度与时间等关键参数的完整个例数据集。本研究总结的系统调查方法，为龙卷灾情调查(包括群发与单个事件)提供了标准化模板，显著提升了工作的准确性与效率。所获得的高精度灾情信息，为构建高质量的龙卷数据库、精准检验监测预警效果提供了不可或缺的实证依据。更多还原

【Abstract】 Based on the tornado outbreak that occurred in four provinces of Eastern China on 17 September 2024, this study develops and implements a systematic damage survey methodology that integrates multi-source information and characteristic damage patterns. The comprehensive approach encompasses key steps such as the collection and analysis of public reports, itinerary planning, on-site investigation techniques, the application of unmanned aerial vehicle(UAV) aerial photography, and detailed damage analysis and assessment. Applying this methodology, the survey team efficiently completes field verification of 10 publicly reported sites within three days. The investigation confirms that 9 of these sites are impacted by 11 independent tornadoes, including 8 rated EF1 and 3 rated EF0, and successfully identifies one false report. The study systematically elaborates on techniques for determining tornado movement direction based on key ground damage features, such as strip damage traces, convergent fallen trees, and narrow passages of damage. Building on this, a standardised UAV aerial survey workflow is established, effectively overcoming common investigation challenges like the subtle ground traces of weak tornadoes and the limitations of radar detection. By synthesising multi-source information, including funnel witness videos, ground damage characteristics, and radar speed paired data, this study realises accurate identification and intensity evaluation of all tornadoes. The result of this process is the construction of a complete case data set, which includes key parameters such as the length, width, intensity, and occurrence time of the disaster path. The systematic survey methodology summarised in this study provides a standardised framework for tornado damage investigation, applicable to both outbreak events and individual tornadoes, and proves to significantly enhance the accuracy and efficiency of the workflow. The high-precision damage information obtained through this framework provides indispensable empirical evidence for constructing a high-quality tornado database and for the precise evaluation of monitoring and warning capabilities.更多还原