Tropical cyclone risk assessment for China at the provincial level based on clustering analysis

• Main Authors: Lin Cai, Yingbing Li, Min Chen, Zixin Zou
• Format: Article
• Language: English
• Published: Taylor & Francis Group 2020-01-01
• Series: Geomatics, Natural Hazards & Risk
• Subjects: tropical cyclones; improved st-dbscan; clustering analysis; risk assessment
• Online Access: http://dx.doi.org/10.1080/19475705.2020.1753823
• ISSN: 1947-5705; 1947-5713

Accurate risk assessment is essential to help allocate the resource for disaster relief and make the evacuation decision when hit by a Tropical Cyclone (TC). The integrated analysis of the hazard-bearing body and the intrinsic characteristics of TCs are promising to provide a clear pattern of the caused risks. However, in the literature, the origin of a TC is not considered in risk assessment. In this article, we propose a risk assessment method aided the analysis of the origins of TCs. Specifically, in order to obtain the knowledge of potential risk, this method first clustered the origins of TCs by an improved ST-DBSCAN algorithm and then make an integrated analysis by combining the analysis results with the hazard-bearing body. A case study is performed to verify the performance of the proposed method. In the study, 1760 TCs generated in the Western North Pacific from 1949 to 2016 are investigated. Results show that the improved ST-DBSCAN can identify regions where TCs generated there have stronger intensities and are more likely to make landfall, and the provincial risks are distinct among TCs generated from different areas. Analysis of annual risk changes in provinces is performed based on the results. These results are valuable for both TC disaster prevention and mitigation.