基于多源信息技术的城市洪涝灾害监测

打开文本图片集

中图分类号：TP79

文献标志码：A

doi：10.3969/j.issn.1000-1379.2026.05.009

Urban Flood Disaster Monitoring Based on Multi-Source Information Technology

SHEN Bingbing， YAN Meichun， WANG Yalu

（College of Geography and Remote Sensing， Hohai University， Nanjing ， China）

Abstract： Flood monitoring is crucial for disaster prevention， reduction and urban sustainability. Taking the “23 ·7” Baoding City center flood event as an example， this study utilized multi-source data， such as volunteer geographic information （VGI）， remote sensing images and precipitation products. Methods such as Python web scraping and random forest classification were employed to extract and comprehensively analyze multi-source information of flood disasters at different temporal scales. The results show that： a） The number of VGI waterlogging points decreases over time， with a primary distribution in the densely populated central urban area. Spatially， these points are distributed from high-density to low-density areas and from the urban center to the periphery， primarily along roads and waterways. This distribution reflects the public’s increased concern for transportation and safety， while also indicating a delayed response from the public. b） The flooded areas extracted from SAR images are sparsely distributed， clearly visible around large water bodies. Different degrees of flooding can be observed in the hotspot areas of waterlogging points， highlighting the collaborative role of remote sensing data and social media data in flood information extraction. c） Cumulative precipitation has a certain impact on the extent of urban flooding， while factors such as water engineering operations also affect flood responses. d） The significantly affected land types by flooding are crops， built areas and expanding water bodies.

Key words： urban flood; machine learning; SAR image; volunteer geographic information; Baoding City

0 引言

洪涝灾害是我国最突出的自然灾害，造成巨大的人员伤亡和经济损失 [1-2] ，尤其是在人口密集且不透水地面比例较高的城市地区，威胁城市发展 [3] 。（剩余10381字）