A Sequentially Coupled Catchment‐Scale Numerical Model for Snowmelt‐Induced Soil Slope Instabilities

Abstract

The frequency of snowmelt-induced soil slope instabilities is increasing in some seasonally cold regions because of climate change. Reliable hazard assessment and risk mitigation of snowmelt-induced landslides require physically-based prediction models. However, existing models either apply only at the slope scale or assume precipitation as the sole landslide trigger. In doing so, they neglect the complexity and coupled nature of the thermo-hydro-mechanical processes leading to slope instability in seasonally cold regions (such as snow accumulation and melting, infiltration and surface runoff, soil saturation, pore water pressure buildup and dissipation). Here, we present a spatially distributed and sequentially coupled numerical model to simulate snowmelt-induced slope instabilities at the catchment scale. The model accounts for temperature-dependent changes in the soil hydraulic behavior related to changes in water state by means of a routine implemented in a geographic information system. We verified the performance of the model using a case study of spring snowmelt-induced soil slope failures that occurred after the 2004 Mid-Niigata earthquake in Japan. Considering limitations and simplifications, the model was able to predict the triggering condition, magnitude, and spatial distribution of the snowmelt-induced landslides with a satisfactory degree of accuracy. We believe that the robustness and simplicity of our numerical approach make it suitable for implementation in early warning systems.Plain Language Summary Climate change is responsible for an increasing number of natural disasters in seasonally cold regions, where heavier snowfalls and sudden snowmelts can cause destructive landslides on hill slopes. These landslides are not easy to study because the mechanisms that trigger them are different from those of shallow landslides occurring in response to rainfall. Moreover, advanced models that incorporate the complex relations among rain, snow, temperature, soil freezing, melting, and slope stability at a detailed scale are still cumbersome and very demanding in terms of computational power. For these reasons, evaluating the risk to people and assets in seasonally cold regions is not straightforward, and existing models tend to underestimate it. In this paper, we propose a novel modeling strategy to study the stability of soil slopes under the influence of both snowmelt and rainfall at a regional scale. First, we explain the governing mechanisms of snowmelt-induced landslides; then, we present our strategy along with detailed explanations of the equations we use. We test our model by predicting a known case study (the snowmelt-induced landslides of 2005 in Niigata, Japan) and obtain a reasonably good result. This simple strategy could be applied in an early warning system for snowmelt-induced landslides.

Thermal Remote Sensing from UAVs: A Review on Methods in Coastal Cliffs Prone to Landslides

“…A method based on the analysis of a series of thermal images (thermograms or thermographic images) acquired during rock-cliff cooling at night-time has been proposed by Zaragoza et al [156] and implemented in MATLAB-MathWorks Inc. environment [166]. The authors motivated the data acquisition during rock cooling as direct sun heating would have induced high disturbance, with significant noises due to the relative position between the source and the facets of the surface.…”

Surface temperature controls the pattern of post-earthquake landslide activity

“…Climatic trends also are considered via changes in precipitation and evapotranspiration 12 – 15 , while temperature oscillations and trends are not accounted for explicitly (as model covariates) 16 , 17 . That is, current LSM approaches do not consider temperature as a possible direct driver of slope instability, even under climate change scenarios, except when discriminating between liquid and solid precipitation 18 or evaluating the depth of the active layer in cold climates.…”

“…Upscaling physically-based models to entire catchments is difficult as simplifications in boundary conditions, parameters, and models are often necessary [40][41][42] . Fully-coupled thermo-hydro-mechanical models have not been implemented at this scale, although partial or sequential couplings have been attempted 18,43 . Currently, geostatistical modeling is the only viable tool for evaluating the non-straightforward (yet possibly direct) effect of temperature on landslide patterns at the catchment scale.…”

A Physically Based Model for the Sequential Evolution Analysis of Rainfall‐Induced Shallow Landslides in a Catchment

“…To account for the different dynamics of surface and subsurface flow, more rigorous models, such as the coupled models combining two or three-dimensional Richards equation solvers and infinite slope stability analysis, have been developed (e.g., Camporese et al, 2010;Rigon et al, 2006;Šimůnek et al, 2016), which often require more computational power. However, the traditional FOS-based model can still be considerably improved; for example, it still cannot reproduce the destruction process of landslides and the force interaction between soil columns (Subramanian et al, 2020). Lehmann and Or (2012) embedded mechanical thresholds resembling concepts of self-organized criticality (SOC) into a hydro-mechanical framework, which simulated a landslide caused by rainfall on a hillslope composed of many interacting soil pillars, and successfully realized the progressive failure process of hillside soil pillars.…”