基于三维聚类识别长江流域连续性极端降水事件时空演变特征
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

湖北省自然科学基金(2024AFD212); 国家自然科学基金项目(42501042,42401030,52109058)


Identifying the spatiotemporal evolution of continuous extreme precipitation events in the Yangtze River Basin based on three-dimensional clustering
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 文章评论
    摘要:

    在全球气候变暖背景下,极端降水事件在范围、强度和频率方面均呈显著上升趋势,严重威胁区域防灾减灾与生态系统的稳定性。当前对极端降水事件识别的研究多从时间或者空间单一角度进行分析,难以全面刻画事件的时间持续性与空间扩展性。以长江流域为研究区,基于1961-2022年逐日0.1°分辨率的栅格降水数据,采用三维聚类方法(经度-纬度-时间)识别流域内连续性极端降水事件,提取其持续时间、影响面积、总强度、平均强度和中心位置等关键特征,系统分析其时空演变规律。研究结果显示:大多数连续性极端降水事件持续时间为2-3d,占总降水事件的90%以上,且这些事件影响面积和强度有限。时间上,连续性极端降水事件的影响面积和总强度均表现出一定的年际波动特征,20世纪90年代之后事件的持续时间出现了明显增加,且总强度较大的极端降水事件发生得更加频繁,尤其是年累加总强度显著增加,说明连续性极端降水事件的累积效应不断增强,潜在致灾风险上升;空间上,连续性极端降水事件在长江流域呈现出明显的区域差异性,具体表现为流域中下游易发生持续时间长、平均强度大的极端降水事件,而总强度大的连续性极端降水事件集中发生在中下游和四川盆地。三维聚类方法实现了对极端降水事件"过程性"和"整体性"的刻画,有助于更好地了解连续性极端降水的时空动态演变,为区域洪涝灾害防范、生态系统稳定性维护及气候变化适应策略提供科学参考。

    Abstract:

    Under the context of global warming, the characteristics of extreme precipitation events have undergone significant changes, with notable increases in spatial extent, intensity, and frequency. These changes pose serious challenges to regional disaster prevention and ecosystem stability. However, existing studies on the identification of extreme precipitation events have predominantly focused on either the temporal dimension or the spatial dimension alone. Such one-dimensional analyses are insufficient to comprehensively characterize both the temporal persistence and spatial expansion of extreme precipitation processes, limiting the comprehensive understanding of their full-dimensional evolution. To address this limitation, this study focused on the Yangtze River Basin, a region highly susceptible to flood hazards, and employed high-resolution gridded daily precipitation data at 0.1° spatial resolution for the period 1961-2022. A three-dimensional clustering algorithm that integrates longitude, latitude, and time dimensions was applied to identify continuous extreme precipitation events across the basin. For each identified event, key characteristics including duration, spatial extent, total intensity, average intensity, and centroid location were extracted to systematically investigate their spatiotemporal evolution patterns. The results revealed that the majority of continuous extreme precipitation events persist for 2-3 days, accounting for more than 90% of the total identified events. These events generally exhibited relatively limited spatial extent and lower intensity. From a temporal perspective, both the spatial extent and total intensity of these events exhibited certain interannual variability. Meanwhile a distinct shift was observed beginning in the 1990s, characterized by a marked increase in event duration and a higher frequency of events with greater total intensity. In particular, the annual cumulative total intensity of these events exhibited a significant upward trend, indicating a strengthening cumulative effect of continuous extreme precipitation and a consequent elevation in potential flood disaster risks. From a spatial perspective, continuous extreme precipitation events exhibited pronounced regional heterogeneity across the Yangtze River Basin. Specifically, the middle and lower reaches of the basin were more prone to experiencing long-duration events with high average intensity, whereas events characterized by large total intensity were mainly concentrated in the middle and lower reaches and in the Sichuan Basin. By integrating spatial and temporal dimensions within a unified analytical framework, the three-dimensional clustering method effectively captures the process-oriented and holistic characteristics of extreme precipitation events. This approach enables a more comprehensive understanding of the spatiotemporal dynamics of continuous extreme precipitation and provides valuable scientific support for regional flood risk management, ecosystem stability maintenance, and the development of climate change adaptation strategies.

    参考文献
    相似文献
    引证文献
引用本文

喻丹,董晓华,张特,王亦洲,魏冲,黄俊雄,李忠安,谈新,彭涛,刘冀.基于三维聚类识别长江流域连续性极端降水事件时空演变特征.生态学报,2026,46(18):9792~9804

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数: