Loading [MathJax]/jax/output/HTML-CSS/fonts/TeX/fontdata.js
Preprints
https://doi.org/10.5194/egusphere-2025-77
https://doi.org/10.5194/egusphere-2025-77
28 Jan 2025
 | 28 Jan 2025
Status: this preprint is open for discussion and under review for Natural Hazards and Earth System Sciences (NHESS).

Identification of nighttime urban flood inundation extent using deep learning

Jiaquan Wan, Xing Wang, Yannian Cheng, Cuiyan Zhang, Yufang Shen, Fengchang Xue, Tao Yang, and Quan J. Wang

Abstract. With the acceleration of urbanization, the disaster of urban flooding has had a serious impact on urban socio-economic activities and has become one of the important factors restricting social development in China. Accurate and timely identification of urban flooding extents is crucial for decision-making. Traditional remote sensing technologies are often limited by environmental factors, making them less suitable for application in complex urban terrains. The development of emerging technologies and the increase in urbanisation have led to a significant increase in the number of surveillance devices within cities, while the development of deep learning techniques has led to their widespread application in various fields. Deep learning methods using video images as a data source provide a new technical methods for intra-urban waterlogging recognition. However, current research mainly focuses on waterlogging recognition in daytime scenes, often ignoring nighttime, a time of high waterlogging incidence. To address these challenges faced by flooding recognition in the nighttime, this study proposes a deep learning model—NWseg—to achieve the recognition of the extent of waterlogging at night. Initially, we constructed a dataset of 4,000 images of nighttime urban flooding. Subsequently, MobileNetV2 and Resnet101 networks were used to replace the DeepLabv3+ backbone network and compared with the NWseg model. Next, the NWseg model is compared with ResNet50-FCN, LRASPP and U-Net models to evaluate the performance of different models in nighttime urban flooding identification. Finally, the applicability and performance differences of each model in specific environments were verified. In conclusion, this study successfully demonstrates the effectiveness of the NWseg model for nighttime urban flooding identification and provides new insights for nighttime urban flooding identification.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.
Share
Download
Short summary
Urban flooding is a growing issue in cities, often disrupting daily life, especially at night...
Share