Scholarlink

<p>Deep-seated landslides are affected by slow and progressive deformation with successive acceleration/deceleration periods controlled by hydro-meteorological factors and slope structures. Simulating groundwater pressures in deep-seated landslides is critical to understanding their mechanisms, forecasting their evolution in time, and estimating landslide hazards. Validating complex deterministic physically based models for deep-seated landslides is challenging in terms of multi-parameter observations and the identification of the physical properties of the media. Using simpler empirical tank models, allowing the simulation of water mass balance and fluxes, is an alternative to simulating their behavior. Tank models must conceptualize the main water reservoirs and use a long history of monitoring observations (rainfall-groundwater levels) to train the model parameters. This study aims at modelling the hydrological dynamics of two deep-seated landslides in fractured rocks (Séchilienne, Mont de La Saxe) with the KarstMod open-source library. Analyzing 6 to 9&#xa0;years of monitoring data allows for the models’ calibration and validation. Two model structures are identified: the Séchilienne landslide’s model, which consists of two tanks and a direct flux to the outlet representing a bimodal flow, and the Mont de La Saxe landslide’s model, which integrates a snow routine and consists of two and a flux feeding a reservoir, representing unimodal flow. The models are validated over 3 and 2&#xa0;years, respectively. They allow the prediction of the discharge and groundwater levels over time, reproducing the slope hydrology in the past before the setup of the monitoring system, thus providing insights into identifying hydrological thresholds controlling landslide motion.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hydrogeological forecasting of deep-seated landslides dynamics: structure and sensitivity of tank models

  • 期刊论文

  • 发表日期：

    2025年2月26日

查看全文
  • Olivier Béjean-Maillard,
  • Catherine Bertrand,
  • Jean-Philippe Malet,
  • Guillaume Cinkus,
  • Pierre Nevers,
  • Joshua Ducasse,
  • Davide Bertolo,
  • Patrick Thuegaz

摘要

Deep-seated landslides are affected by slow and progressive deformation with successive acceleration/deceleration periods controlled by hydro-meteorological factors and slope structures. Simulating groundwater pressures in deep-seated landslides is critical to understanding their mechanisms, forecasting their evolution in time, and estimating landslide hazards. Validating complex deterministic physically based models for deep-seated landslides is challenging in terms of multi-parameter observations and the identification of the physical properties of the media. Using simpler empirical tank models, allowing the simulation of water mass balance and fluxes, is an alternative to simulating their behavior. Tank models must conceptualize the main water reservoirs and use a long history of monitoring observations (rainfall-groundwater levels) to train the model parameters. This study aims at modelling the hydrological dynamics of two deep-seated landslides in fractured rocks (Séchilienne, Mont de La Saxe) with the KarstMod open-source library. Analyzing 6 to 9 years of monitoring data allows for the models’ calibration and validation. Two model structures are identified: the Séchilienne landslide’s model, which consists of two tanks and a direct flux to the outlet representing a bimodal flow, and the Mont de La Saxe landslide’s model, which integrates a snow routine and consists of two and a flux feeding a reservoir, representing unimodal flow. The models are validated over 3 and 2 years, respectively. They allow the prediction of the discharge and groundwater levels over time, reproducing the slope hydrology in the past before the setup of the monitoring system, thus providing insights into identifying hydrological thresholds controlling landslide motion.