台风扰动下江苏海岛生态环境动态特征
Dynamic characteristics of Jiangsu island ecological environment under typhoon disturbance
丁慧文 1孙佳龙 2陈广欣1
作者信息
- 1. 江苏海洋大学海洋技术与测绘学院，江苏连云港 222005
- 2. 江苏海洋大学海洋技术与测绘学院，江苏连云港 222005;江苏省海洋遥感工程研究中心，江苏连云港 222005;江苏省高校海洋气象防灾减灾重点实验室（江苏海洋大学）, 江苏连云港 222005
- 折叠
摘要
针对台风扰动下海岛生态系统动态响应机制的问题，本文以江苏9个典型海岛为对象，构建涵盖生态本底、人类干扰、治理响应和风险胁迫的14项指标体系，提出融合分段贝叶斯网络与时变熵权法的动态评价模型，基于2014-2024年多源数据，以2018年台风"温比亚"为断点开展研究。结果表明：台风后极端天气发生次数指标权重显著上升，升幅达251.0549%，而归一化植被指数（normalized difference vegetation index，NDVI）植被覆盖率等本底指标权重保持稳定；依据灾后恢复特征，岛屿可分为高韧性恢复型（生态综合得分提升10.10%~14.84%）、稳定适应型（生态综合得分分值波动<7%）和敏感衰退型（生态综合得分下降>3%）三类；双重差分模型（difference-in-differences，DID）验证了台风对高暴露岛屿的显著负面影响。研究建议实施"一岛一策"的差异化治理，通过风险预警与生态修复提升海岛气候韧性，为海岛生态系统适应性管理提供科学依据。
Abstract
To address the problem of understanding the dynamic response mechanisms of island ecosystems under typhoon disturbance, this study focuses on nine typical islands in Jiangsu Province. An indicator system comprising 14 items, covering ecological background, human disturbance, governance response, and risk stress, was constructed. A dynamic evaluation model integrating a segmented Bayesian network and a time-varying entropy weight method was proposed. Based on multi-source data from 2014 to 2024, and using the 2018 Typhoon Rumbia as a breakpoint, the research was conducted. The results indicate that after the typhoon, the weight of extreme weather indicators increased significantly, by up to 251.0549%, while the weights of background indicators such as vegetation coverage remained stable. Based on post-disaster recovery characteristics, the islands can be categorized into three types: high-resilience type (score increase of 10.10%~14.84%), stable-adaptation type (score fluctuation < 7%), and sensitive-declining type (score decrease > 3%). A difference-in-differences (DID) model verified the significant negative impact of the typhoon on highly exposed islands. The study recommends implementing differentiated governance strategies tailored to individual islands ("one island, one policy") to enhance island climate resilience through risk early warning and ecological restoration, providing a scientific basis for the adaptive management of island ecosystems.关键词
时变熵权法/贝叶斯网络/海岛生态质量/多源遥感/DPSIR 框架Key words
time-varying entropy weight method/Bayesian network/island ecological quality/multi-source remote sensing/DPSIR framework引用本文复制引用
丁慧文,孙佳龙,陈广欣.台风扰动下江苏海岛生态环境动态特征[EB/OL].(2026-05-14)[2026-05-15].http://www.paper.edu.cn/releasepaper/content/202605-52.学科分类
环境科学技术现状/灾害、灾害防治/环境管理
评论