Modeling Jamming Caused by Debris Flows Through a Series of Cascading Structures
Abstract
Debris flows are a hazard present in many mountain regions. They typically occur when steep mountain creeks flood, transporting sediment, woody debris, and often large boulders. Large boulders can cause jamming through constrictions during debris flow events, affecting the outflow rate of the debris flow, and causing sediment deposition. In 2022, Piton et al. developed a model to possibilistically simulate boulder jamming in a debris basin, and its effects on multi-phase flow. The Piton et al. (2022, Journal of Geophysical Research, DOI: 10.1029/2021JF006447) model uses simple hydraulic equations, basic geometry, and stochastic generation of boulders to simulate jamming. The possibilistic nature of the model allows it to encompass the uncertainty associated with the simulation of debris flows through uncertainty propagation. In this study, we present an extension to the Piton et al. (2022) framework, wherein jamming can be possibilistically simulated through a series of debris basins or bridges (previously, the Piton et al. (2022) model was only able to simulate jamming through a single debris basin). The updated framework is able to route the event hydrograph between consecutive structures and can either route stochastically generated boulders directly to the downstream structure or regenerate boulders if deposition/erosion (i.e. "mixing") is expected between consecutive structures. The model framework is freely available through an online interface at https://platrisk.ige.inrae.fr. The code is open source and available on GitHub where it can be downloaded for R. The updated framework was used to simulate the Torrent du Saint-Martin, a steep creek within the Maurienne Valley, France. The input parameters used to simulate jamming within the Torrent du Saint-Martin were measured, researched, or estimated using a variety of methods. Two simulations were conducted: a re-creation of the June 2005 event (a historical debris flow event that occurred along the Torrent du Saint-Martin) to compare the results of the updated framework against a known historical event, exploring the validity of the approach; and a parametric analysis to assess the monotony of the input variables and the model's sensitivity. The results of the simulation of the 2005 event aligned with the historical records. The simulated jamming height and deposited debris flow volume upstream of the modeled structures aligned with the historical records of the debris flow. The results of the parametric analysis also aligned with expectations. The simulated monotonies of the input variables were consistent with expected debris flow behavior and were added to the framework to reduce model runtime. Most model inputs (with the exception of event volume and peak inflow) had a minor impact on the model results, varying the outflow volume and peak outflow by 15% or less, indicating that the model is robust in the face of uncertainty. Overall, the results of the analyses indicate that the model can be effectively used to simulate jamming caused by a debris flow through a series of debris basins or bridges.