Monitoring & inspection |
Smarter monitoring for China’s dams
Monitoring seismicity in the Yalong River Basin and using close range photogrammetry have been the focus of recent research efforts in China.
separates earthquake-induced plastic deformation from other components. It also updates dam– foundation–abutment rock mass stress and deformation fields online to help identify damage.
● By structuring monitoring, inspection and structural calculation results into a three-axis early-warning matrix, the framework achieves multi-source information fusion and graded post-earthquake safety assessment.
Photogrammetry Recent advances in computer vision and sensor
Damage after the earthquake in Wenchuan, China. Image credit: Rick Siu/
Shutterstock.com
RICH IN HYDROPOWER RESOURCES, Southwest China has become a magnet for dam construction even though the region lies in one of the country’s major earthquake zones, characterised by frequent seismicity and a high probability of strong events. Past events include the Wenchuan earthquake on 12 May 2008, which caused structural damage to the Zipingpu Dam, and the 5 September 2022 Luding earthquake, which induced a plastic deformation increment at the Dagangshan dam. As Liu et al write in Frontiers in Earth Science, rapid post-earthquake assessment of large dams is essential in seismically active mountain regions. However, compared with analysis and design, research on post-earthquake dam safety monitoring and rapid structural assessment remains relatively limited, while operational monitoring and early-warning technology under earthquake conditions is still a blind spot.
References
Development of an online seismic disaster monitoring system in Yalong river by Jian Liu, Peng Zhang, Chen Zhang, Qing-xiang Meng, Zijie He and Tao Feng. Frontiers in Earth Science. 14:1747819. doi: 10.3389/ feart.2026.1747819
Research on Dam Inspection Method Based on Close-range Photogrammetry by Shu Yang and Feng Yin. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume X-1/W2-2025. https://
doi.org/10.5194/isprs-annals-X- 1-W2-2025-187-2025
With this in mind, a Chinese collaborative research team developed an online dam-safety monitoring framework for cascade hydropower stations in the Yalong River Basin. The basin has an active seismotectonic setting with steep gorges and unstable slopes, while dam foundations and abutment rock masses are commonly cut by faults, joint sets and bedding or foliation. These can influence deformability, potential sliding and dam–foundation– abutment response under seismic loading. This context motivated a basin-wide operational framework linking seismic triggering, monitoring interpretation and numerical diagnosis, while recognising that site-specific geological conditions affect the interpretation and applicability of system outputs. The study focused on earthquake-resistant dam safety management, integrating seismic monitoring, deformation modelling and numerical simulation within a unified early- warning scheme. The main conclusions were: ● The framework provides dual-source seismic triggering and rapid post-earthquake data acquisition, enabling automated intensified patrols and allowing cascade hydropower stations to obtain key safety monitoring information within minutes.
● The elastoplastic deformation monitoring model establishes a monitoring–simulation chain that
28 | September 2026 |
www.waterpowermagazine.com
technologies have increased the use of photogrammetry for structural health monitoring due to its non-contact nature, efficiency and precision. A novel dam inspection method based on close-range photogrammetry aims to address limitations of traditional dam safety monitoring, including low efficiency, high cost and limited coverage. The study conducted experiments on three types of dams using high-resolution cameras mounted on drones for close-range photogrammetry, combined with computer vision and deep learning algorithms. The inspection system comprises: high-resolution image acquisition scheme for dam surfaces; automatic defect identification algorithm based on deep learning; 3D model reconstruction and deformation analysis technology; and engineering application verification and precision evaluation. The proposed method achieved crack detection accuracy of ±0.1mm and deformation monitoring accuracy of ±1.2mm. Compared with traditional methods, efficiency improved by five to eight times and costs were reduced by more than 60%. The proposed MFFN deep learning algorithm achieved an accuracy rate of over 95% in identifying typical dam defects, significantly reducing manual interpretation. The method also showed applicability to concrete, earth and rockfill, and arch dams.. Future research could include:
● Combining data from InSAR, LiDAR, and other sources to enhance the dimensions of monitoring.
● Developing real-time processing systems based on UAVs.
● Using digital twins for constructing a health monitoring platform for the entire lifecycle of dams.
● Implementing AI-based adaptive inspection strategies for anomalies. Authors Shu Yang and Feng Yin believe the approach
provides an efficient, precise and cost-effective solution for dam safety monitoring and could support intelligent inspection across multiple dam types. Looking ahead, they add that close-range photogrammetry could become increasingly integrated with automated monitoring systems and IoT platforms, supporting improvements in dam safety management in China..
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45
