A multi-platform, open-source, and quantitative remote sensing framework for dam-related hazard investigation: Insights into the 2020 Sardoba dam collapse

Abstract

Monitoring dam deformation and reservoir conditions plays an important role in routine dam safety assessment. However, maintaining a real-time dam monitoring system and conducting frequent site surveys come with high costs, which may hinder the detection and mitigation of potentially hazardous dam conditions or incident investigations. In view of inaccessibility of in situ data for the scientific community and the need for rapid dam failure hazard investigation, we propose an end-to-end framework that relies on multiple and open-source remote sensing data for the dam-related hazard investigation (RSDHI). The proposed RSDHI framework includes three modules that are capable of monitoring the post-construction deformation status of the dam and the reservoir and provide the first-order and quantitative examination of the hazard causality based on numerical models. We apply and validate the RSDHI framework to a case study of the 2020 Sardoba dam failure in Uzbekistan. We show that the Sardoba Dam experienced continuing subsidence and a local ∼ 4.7 cm differential settlement near the breached section. We reveal that the secondary consolidation controls post-construction deformation. The failure was likely related to the compound effect of transverse structural cracks that resulted from the differential settlement and water loading, yet the exact reason for its failure remains unknown. This study demonstrates that the RSDHI framework can combine multi-sensor remote sensing observations and numerical modeling to provide a complete and up-to-date status report of dam conditions, thereby providing insights into potential instability of dams around the world and enabling rapid investigation of their failures even when no in situ data is available.