Understanding multiscale drivers of natural hazards, cascading failures, and risk management strategies within a multisector system

Deleterious impacts of rapid unplanned anthropogenic disturbances have been compounded by climate change globally. This phenomenon is particularly prominent in the high mountain regions that have suffered a string of cascading hazard-related disasters (CHDs). Recent catastrophic events (e.g., 2013 Uttarakhand Flood, 2021 Chamoli Landslide, and 2021 Melamchi Debris/Flood) have highlighted the need to better understand the complex interactions among human, natural, and engineered systems to inform the design of disaster management strategies. It is crucial to rethink disaster management as a multisystem-connected problem. In such a deeply interconnected system, it is essential to build a systematic framework to reveal linkages and identify spatially and temporally varying risk probabilities. We develop data-driven models that integrate existing hydroclimatic models (e.g., glacial lake outburst flood, landslide, and flood) and data (e.g., NASA Earth Observations) with non-traditional data streams (e.g., Citizen Science and expert knowledge) to investigate connections that lead to CHDs.

Our modeling framework synergistically integrates models and data from different systems using a Bayesian network. The framework will serve as an operational system-of-systems model for the high mountain region that can formalize how Citizen Science and expert knowledge may be utilized with existing models for managing CHDs. Here the experts refer to everyone involved in decision-making, including academic researchers, public agency researchers, policymakers, and managers on the ground. We propose that a cyberinfrastructure should be developed that integrates all data streams and model resources necessary to understand the spatially and temporally varying risks. The cyberinfrastructure will facilitate ‘what-if’ type analysis to understand system dynamics and sensitivity to perturbations that may be used to design mitigation strategies.

 

Case Study

Specifically, we choose Nepal Himalaya where natural hazards and cascading failure are a major concern. The region is characterized by extreme elevation gradient, young and fragile geology, extreme seasonal and spatial variation in rainfall, and diverse human impacts. One hazard often triggers another hazard in the region, leading to cascading disaster. Also, a seemingly non-hazardous series of average events can trigger a chain of events over a long or short time-scale with disastrous consequences. Knowledge and understanding of these connections are essential for planning mitigation measures and improving hazards predictions in the region.