From near-miss to collapse: diagnosing temporal governance gaps in landslide-dammed lake risk management

Landslide-dammed lakes are produced by mass-movement processes and may remain apparently stable before rapid breach or collapse. This study examines the Matai’an Creek landslide-dammed lake in eastern Taiwan to diagnose why recognized hazard signals did not translate into timely structural action. This study develops a temporal diagnostic framework that combines documentary process tracing, case-policy evaluation, phase-based coding of 97 disaster risk reduction actions, and an Escalation Index (EI) to evaluate governance depth across four phases: hazard formation, monitoring and coordination, collapse, and post-collapse adjustment. The analysis clarifies the geomorphological setting, data sources, boundary-delineation workflow, and uncertainty limits, while distinguishing documented evidence from interpretive governance inference. Results show that pre-collapse actions were dominated by monitoring and coordination (EI = 1.36–1.61), whereas structural mitigation and trigger-based intervention emerged mainly after collapse (EI = 3.38–3.42). The findings indicate a temporal governance gap between signal recognition and escalation activation, and demonstrate how near-miss periods can be used as policy stress tests for anticipatory landslide-dammed lake risk management. A temporal diagnostic framework links warning signals to governance escalation. Ninety-seven DRR actions are coded by phase and escalation depth. The Escalation Index quantifies the shift from procedural to structural action. Monte Carlo tests confirm a robust pre-/post-collapse escalation contrast. Near-miss periods reveal missed trigger windows for anticipatory governance.