An Operational Rockfall Monitoring Framework for Hazard Management: A Case Study of the Daman Slope, Taiwan

Rockfall hazards pose a persistent threat to mountain road safety, particularly along high-risk corridors in regions affected by frequent earthquakes and intense rainfall, where sudden slope failures directly constrain long-term road operations and place road users at risk. In many such corridors, short-term engineering mitigation is not feasible, yet road operations must be sustained over extended periods, making disaster prevention reliant on monitoring, warning, and operational control rather than structural solutions. This study presents the Daman slope, located at 49.8 km along Provincial Highway No. 7 in Taoyuan, Taiwan, as a representative case demonstrating how slope monitoring has evolved into a practical disaster prevention system under these constraints. Early monitoring efforts focused on compiling an event catalog and evaluating rockfall occurrence sensitivity derived from a microtremor system to support operational decisions, such as adjusting traffic access frequency to reduce exposure during periods of elevated activity. While this sensitivity-based approach provided an initial framework for risk management, subsequent experience showed that it was insufficient for operational decision-making when hazards were triggered by earthquakes and intense rainfall, as strong seismic motions exceeded the effective range of the microtremor-based monitoring system, while rainfall-induced conditions were associated with elevated noise levels that reduced signal reliability. Such events are characterized by abrupt onset and severe consequences, particularly when rockfalls occur during active traffic operations, leaving little opportunity for advance intervention. The limitations of prediction became evident during the 3 April 2024 Mw 7.2 Hualien earthquake, when strong ground motion triggered multiple rockfalls during seismic shaking without identifiable precursory signals; similar challenges were also observed for rainfall-related rockfalls, reinforcing the recognition that such hazards cannot be reliably forecast using sensitivity indicators alone. As a result, the monitoring strategy transitioned from an analysis focused on prediction toward a framework centered on warning and disaster prevention. The system was expanded to integrate ground motion and rainfall observations in real time, with an emphasis on identifying hazardous conditions that require immediate operational response. A standardized operating procedure has been established to ensure that monitoring information is consistently translated into warning displays and traffic management actions at the site. In current practice, warning levels displayed in the early morning are determined based on monitoring records from the preceding night, while daytime operations generally allow full access, with warning signals adjusted dynamically when monitored conditions exceed predefined thresholds. Within this framework, the core function of the system remains focused on rapid hazard recognition and warning issuance based on direct monitoring observations and predefined operational thresholds, while artificial intelligence techniques are applied in post-processing as supportive tools to refine event interpretation and improve the accuracy and consistency of the event catalog. This case highlights how slope monitoring can function as an active disaster prevention mechanism by shifting the emphasis from attempting to predict individual failures to reducing exposure and enhancing road user safety through timely warning and operational control when engineering mitigation is constrained.