Over the past decades, China has experienced severe compound natural disasters, such as extreme rainfalls, which have led to significant losses. In response to the challenges posed by the lack of a clear investigation process and inadequate comprehensiveness in evaluating the natural disaster chains, this study proposes a comprehensive retrospective simulation strategy for emergency investigation and simulation of heavy rainstorm–flash flood–debris flow chain disasters at the county–town level. The primary aim is to avert potential new chain disasters and alleviate subsequent disasters. This study combines emergency investigation efforts with hydrodynamic models to digitally simulate and analyze compound chain disasters triggered by an extreme rainfall event in the Haihe River regional area, specifically Gaoyakou Valley, Liucun Town, Changping District, Beijing, in July 2023, along with potential new disasters in adjacent regions. The findings indicate that the heavy rainstorm chain disaster on “7.29” resulted from a complex interplay of interrelated natural phenomena, including flash floods, debris flows, urban floodings, and river overflows. Hantai Village has experienced flash flood and debris flow events at different times in this area. Should the rainfall volume experienced in Liucun Town be replicated in the Ming Tombs Town area, approximately 6.2 km² of land would be inundated, leading to damages to 458 residences and impacting around 240 ha of agricultural land. The anticipated release of floodwater from the reservoir would lead to significant impacts on downstream residents and roads. Our research can improve the efficacy of emergency investigations and assessments, which in turn can help with the management and reduction of disaster risks at the grassroots level.

Piping is a major factor contributing to river embankment breaches, particularly during flood season in small and medium rivers. To reduce the costs of earth rock embankment inspections, avoid the need for human inspectors and enable the quick and widespread detection of piping hazards, a UAV image-acquisition function was introduced in this study. Through the collection and analysis of thermal infrared and visible (TIR & V) images from several piping field simulation experiments, temperature increases, and diffusion centered on the piping point were discovered, so an automatic algorithm for piping identification was developed to capture this phenomenon. To verify the identification capabilities, the automatic identification algorithm was applied to detect potential piping hazards during the 2022 flooding of the Dingjialiu River, Liaoning, China. The algorithm successfully identified all five piping hazard locations, demonstrating its potential for detecting embankment piping.

Levees are critical for flood defence, but their integrity is threatened by hazards such as piping and seepage, especially during high-water-level periods. Traditional manual inspections for these hazards and associated emergency response elements, such as personnel and assets, are inefficient and often impractical. While UAV-based remote sensing offers a promising alternative, the effective fusion of multi-modal data and the scarcity of labelled data for supervised model training remain significant challenges. To overcome these limitations, this paper reframes levee monitoring as an unsupervised anomaly detection task. We propose a novel, fully automated framework that unifies geophysical hazards and emergency response elements into a single analytical category of “abnormal targets” for comprehensive situational awareness. The framework consists of three key modules: (1) a state-of-the-art registration algorithm to precisely align infrared and visible images; (2) a generative adversarial network to fuse the thermal information from IR images with the textural details from visible images; and (3) an adaptive, unsupervised segmentation module where a mean-shift clustering algorithm, with its hyperparameters automatically tuned by Bayesian optimization, delineates the targets. We validated our framework on a real-world dataset collected from a levee on the Pajiang River, China. The proposed method demonstrates superior performance over all baselines, achieving an Intersection over Union of 0.348 and a macro F1-Score of 0.479. This work provides a practical, training-free solution for comprehensive levee monitoring and demonstrates the synergistic potential of multi-modal fusion and automated machine learning for disaster management.

The analysis of temporal and spatial variability in risk has garnered significant research attention, particularly regarding flash flood disasters in the context of warming and wetting conditions on the Qinghai–Tibetan Plateau. Focusing on the Lhasa River basin, this study develops a framework that integrates geographic information systems and a combined subjective–objective weighting approach to comprehensively assess flash flood risk despite limited observations. This paper investigates the distribution patterns of hazard, vulnerability, and the integrated risk of pluvial flash floods; demonstrates the reliability of the assessment results; and provides mitigation recommendations for disaster risk management at the county level. The results showed a trend in increasing flash flood risk in recent decades compared to the 1990s. Moreover, very-high- and high-risk areas were concentrated in downstream regions with frequent precipitation extremes and anthropogenic activity. From 1991 to 2020, the high to very high-risk areas gradually expanded from central Lhasa to neighbouring counties. This study contributes valuable insights into flash flood risk assessment cand mapping, which are crucial in terms of the protection of life and property in the plateau basin.

Embankment piping and leakage are primary causes of flood control infrastructure failure, accounting for more than 90% of embankment failures worldwide and posing significant threats to public safety and economic stability. Current manual inspection methods are labor-intensive, hazardous, and inadequate for emergency flood season monitoring, while existing automated approaches using thermal infrared imaging face limitations in cost, weather dependency, and deployment flexibility. This study addresses the critical scientific challenge of developing reliable, cost-effective automated detection systems for embankment safety monitoring using Unmanned Aerial Vehicle (UAV)-based visible light imagery. The fundamental problem lies in extracting subtle textural signatures of piping and leakage from complex embankment surface patterns under varying environmental conditions. To solve this challenge, we propose the Embankment-Frequency Network (EmbFreq-Net), a frequency-enhanced deep learning framework that leverages frequency-domain analysis to amplify hazard-related features while suppressing environmental noise. The architecture integrates dynamic frequency-domain feature extraction, multi-scale attention mechanisms, and lightweight design principles to achieve real-time detection capabilities suitable for emergency deployment and edge computing applications. This approach transforms traditional post-processing workflows into an efficient real-time edge computing solution, significantly improving computational efficiency and enabling immediate on-site hazard assessment. Comprehensive evaluations on a specialized embankment hazard dataset demonstrate that EmbFreq-Net achieves 77.68% mAP@0.5, representing a 4.19 percentage point improvement over state-of-the-art methods, while reducing computational requirements by 27.0% (4.6 vs. 6.3 Giga Floating-Point Operations (GFLOPs)) and model parameters by 21.7% (2.02M vs. 2.58M). These results demonstrate the method’s potential for transforming embankment safety monitoring from reactive manual inspection to proactive automated surveillance, thereby contributing to enhanced flood risk management and infrastructure resilience.