Elsevier

CATENA

Volume 201, June 2021, 105213
CATENA

Fast physically-based model for rainfall-induced landslide susceptibility assessment at regional scale

https://doi.org/10.1016/j.catena.2021.105213Get rights and content

Highlights

  • A fast physically-based model for regional landslide susceptibility assessment.

  • Both lateral and vertical flows were considered to calculate the water table.

  • Stochastic approach for the input parameters cohesion and friction angle.

  • Model application by using landslide inventory in Andorra, Pyrenees.

  • Very short computational time (~2 min) to calculate a large area (~479 km2).

Abstract

Rainfall-induced landslides represent an important threat in mountainous areas. Therefore, a physically-based model called “Fast Shallow Landslide Assessment Model” (FSLAM) was developed to calculate large areas (>100 km2) with a high-resolution topography in a very short computational time. FSLAM applies a simplified hydrological model and the infinite slope theory, while the two most sensitive soil properties regarding slope stability (cohesion and friction angle) can be stochastically included. The model has five necessary input raster files including information of soil properties, vegetation, elevation and rainfall. The principal output is the probability of failure (PoF) map. The Principality of Andorra was selected as case study, where FSLAM was successfully applied and validated using the existing landslide inventory. The PoF raster file of Andorra (including 19 million cells) was calculated in only 2 min. Therefore, an accurate calibration of the input parameters was easy, which strongly improved the final outcomes.

Introduction

Landslides represent one of the most significant geomorphological hazards in mountainous regions and characterise an important risk for people and infrastructures (Petley, 2012, Kirschbaum et al., 2015, Froude and Petley, 2018). Especially shallow landslides that are triggered by rainfall exceeding a certain threshold, can cause considerable losses (Caine, 1980, Guzzetti et al., 2007, Hungr et al., 2014). It is therefore a fundamental task for stakeholders and research institutions to perform an adequate landslide susceptibility, hazard and risk assessment (Guzzetti et al., 1999, Crozier, 2005, Fell et al., 2008, Segoni et al., 2020).

The susceptibility analysis identifies the prone areas where landslides can initiate and propagate (Guzzetti et al., 2005, Guzzetti et al., 2006), and is the starting point of each hazard and risk assessment (Corominas et al., 2014). One of the results obtained from this task is the landslide susceptibility map, which is generally used for the management of territories or the land use planning (Fell et al., 2008, Chen et al., 2019). It is evident that approaches applied to landslide susceptibility modelling are highly essential for the reliability of the resulting map, and even a small increment in prediction accuracy may have a large impact.

Numerous models have been developed for the purpose of landslide susceptibility assessment. They can be divided into four main categories: expert-based models, physically-based models, statistical models, and machine learning models (Guzzetti et al., 1999, Huang et al., 2017, Sezer et al., 2017, Broeckx et al., 2018, Reichenbach et al., 2018). Among these models, the expert-based models mainly use the expert opinion to build the explanatory variables. However, their opinions are highly subjective and therefore the analysed process is difficult to be replicated by other users to different areas (Kirschbaum et al., 2016, Hearn and Hart, 2019). The statistical and machine learning models are normally considered as data-driven approaches, both of which focus on the analysis of the landslide influencing factors using past and present landslide datasets (Goetz et al., 2015, Huang and Zhao, 2018, Zêzere et al., 2017). Although in some cases the data-driven approaches have been considered more accurate than other approaches, they ignore the complex physical processes involved in landslide initiation. On the contrary, physically-based models can take geotechnical characteristics of landslides into account, and normally quantify the slope stability by combining the infinite slope stability approach and hydrological assumptions. Hence, one important advantage of physically-based models is to calculate slope stability by using physical properties that control geomorphological processes, which better reflects the mechanism of landslide occurrences. The physically-based technique has been frequently adopted for the susceptibility assessment of shallow landslides and a list of available codes is given in Table 1. Most models use Mohr Coulomb theory to achieve geotechnical modelling, such as SHALSTAB and TRIGRS, while a few models use other theories, including STEP-TRAMM, SCOOPS 3D and R.ROTSTAB. Regarding the hydrological modelling, some models consider both lateral and vertical flows of groundwater (e.g., GEOtop-FS and HIRESSS) but most models only involve one of them (e.g., SINMAP model).

In this study, we developed a new physically-based model, because the quantitative analysis of future climate and vegetation changes was one of the requirements of the code and only physically-based models are able to include different rainfall scenarios and the effect of root strength due to vegetation. However, physically-based models have also limitations like the ones referring to the determination of the soil properties, which generally includes a large uncertainty (e.g. Tofani et al., 2017). In particular, the site-specific data generally is seldom available at regional scale (Carrara et al., 2008). Consequently, the regional-scale application of physical models for landslide susceptibility zonation is increasingly being an operational challenge. Another drawback of physically-based models is the high computational cost, especially when a comprehensive approach for the rainfall infiltration in unsaturated soil is used (e.g. the incorporation of the so-called Richards equation; (Iverson, 2000)). In our new model, the geotechnical component is based on the infinite slope theory, whereas the hydrological model combines the lateral and vertical flows to calculate the water table. The input parameters consider various aspects that have impacts on slope stability, including digital elevation model (DEM), soil properties, vegetation and rainfall. Finally, a stochastic approach is applied for soil properties to include their uncertainty. With the purpose of addressing spatial uncertainty, it would have required including the space correlation.

The main objectives of this work include: (i) the presentation of a novel physically-based model, called Fast Shallow Landslide Assessment Model (FSLAM), (ii) the evaluation of the different input parameters by a sensitivity analysis, and (iii) the application and evaluation of the proposed model at regional scale. Herein, the definition of regional scale stands for areas with an extension of more than 100 km2, like our study area of Andorra, which covers about 470 km2.

Section snippets

General settings

The Principality of Andorra is situated in the middle of the Pyrenees (Fig. 1) and covers a total area of approximately 470 km2. The country has a typical high-mountain morphology with the capital of Andorra la Vella situated at about 1000 m asl. The morphology of Andorra has been strongly shaped by glacial and fluvial processes. The resulting main valleys include the two valleys Valira d’Orient and Valira del Nord that unify at the capital Andorra la Vella to form the Gran Valira valley (Fig. 1

General aspects

The main assumption underlying the FSLAM model is to consider the rainfall as the landslide triggering mechanism. The rainfall effect appears in the soil mechanics as the pore pressure reducing the effective stresses that contribute to destabilize the soil.

Slope stability models mostly involve two sub-models: on one side a mechanical or geotechnical model, and on the other side a hydrological model focussing to the water flow. Regarding the geotechnical model, the infinite slope theory has been

Model verification

To perform the verification of the FSLAM code, a homogenous slope with the size of 100 × 100 m and a constant slope angle of 25° was created. The cell size of the input raster was 10 m. The factor of safety (FS) of each cell in the slope was calculated changing the input parameters of our model. The ranges of the input values and their default values were selected using standard literature and expert criteria (Geotechdata, 2020, USDA, 1986, ECORISQ, 2020) (Table 2). In parallel, the FS-values

Discussion

Complex physically-based models are often used to assess rainfall-induced landslide susceptibility at regional scale, but computational cost is normally high (e.g., Alvioli et al., 2016, Pourghasemi and Rahmati, 2018). Although modern computational hardware has been strongly improved, the computational time can still need several days for a relatively small area at high temporal and spatial resolution (Rossi et al., 2013). This drawback is mainly related to the calculation of the rainfall

Conclusions

The susceptibility assessment of rainfall-induced shallow slope failures at regional scale (>100 km2) is a complex task. In the present study, a novel model called FSLAM was proposed aiming to achieve fast assessment of landslide susceptibility on large areas.

The sensitivity analysis and Pareto plot revealed that the cohesion (Cr and Cs) and the friction angle (φ) are the most important input parameters. Thus, they are included into the model as stochastic input parameters. In contrast, the

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The study was funded by the national research projects called SMuCPhy (BIA 2015-67500-R) and EROSLOP (PID2019-104266RB-I00) granted by the Spain Government and co-funded by AEI/FEDER, UE. Zizheng Guo acknowledges the financial support of China Scholarship Council for his research at UPC BarcelonaTECH, and Fundamental Research Funds for National Universities, China University of Geosciences (Wuhan).

References (87)

  • E.E. Alonso

    Risk analysis of slopes and its application to slopes in Canadian sensitive clays

    Geotechnique

    (1976)
  • M. Alvioli et al.

    Evaluation of the parallel performance of the TRIGRS v2.1 model for rainfall-induced landslides

    PeerJ

    (2016)
  • Andorra Environmental Ministry, 2015. Estudi per a l’elaboració Anyal del balanç hídric del Principat d'andorra...
  • E. Aristizábal et al.

    SHIA_Landslide: a distributed conceptual and physically based model to forecast the temporal and spatial occurrence of shallow landslides triggered by rainfall in tropical and mountainous basins

    Landslides

    (2016)
  • Baeza, C., 1994. Evaluación de las condiciones de rotura y de la movilidad de los deslizamientos superficiales mediante...
  • R.L. Baum et al.

    Estimating the timing and location of shallow rainfall-induced landslides using a model for transient, unsaturated infiltration

    J. Geophys. Res. Earth Surf.

    (2010)
  • Baum, R.L., Savage, W.Z., Godt, J.W., 2008. TRIGRS — A Fortran Program for Transient Rainfall Infiltration and...
  • R.L. Baum et al.

    TRIGRS-A fortran program for transient rainfall infiltration and grid-based regional slope-stability analysis

    (2002)
  • M. Bordoni et al.

    Empirical and physically based thresholds for the occurrence of shallow landslides in a prone area of Northern Italian Apennines

    Water (Switzerland)

    (2019)
  • J. Broeckx et al.

    A data-based landslide susceptibility map of Africa

    Earth-Sci. Rev.

    (2018)
  • N. Caine

    The rainfall intensity-duration control of shallow landslides and debris flows

    Geogr. Ann.

    (1980)
  • L. Cascini et al.

    Spatial and temporal occurrence of rainfall-induced shallow landslides of flow type: a case of Sarno-Quindici, Italy

    Geomorphology

    (2011)
  • CENMA, 2008. Informe sobre l’episodi de precipitacions intenses i torrentada al riu Runer i torrent de Fontanals de l’1...
  • L. Chen et al.

    The influence of land use and land cover change on landslide susceptibility: a case study in Zhushan Town, Xuan’en County (Hubei, China)

    Nat. Hazards Earth Syst. Sci.

    (2019)
  • S.E. Cho

    Effects of spatial variability of soil properties on slope stability

    Eng. Geol.

    (2007)
  • J. Corominas et al.

    Landslide rainfall triggers in the Spanish Eastern Pyrenees

  • J. Corominas et al.

    Recommendations for the quantitative analysis of landslide risk

    Bull. Eng. Geol. Environ.

    (2014)
  • M. Crozier

    Multiple-occurrence regional landslide events in New Zealand: hazard management issues

    Landslides

    (2005)
  • Dietrich, W.E., Montgomery, D.R., 1998. SHALSTAB: a digital terrain model for mapping shallow landslide potential....
  • Digital Climate Atlas of Andorra, 2020. https://www.iea.ad/cartografia/acda-atles-climatic-digital-d-andorra (accessed...
  • ECORISQ, 2020. https://www.ecorisq.org/docs/USCS_soilclasses.pdf (accessed 1 February...
  • H. El-Ramly et al.

    Probabilistic slope stability analysis for practice

    Can. Geotech. J.

    (2002)
  • R. Fell et al.

    Guidelines for landslide susceptibility, hazard and risk zoning for land use planning

    Eng. Geol.

    (2008)
  • M.J. Froude et al.

    Global fatal landslide occurrence from 2004 to 2016

    Nat. Hazards Earth Syst. Sci.

    (2018)
  • W.R. Gardner

    Some steady-state solutions of the unsaturated moisture flow equation with application to evaporation from a water table

    Soil Sci.

    (1958)
  • Gelonch Roca, R., 2014. Elaboración de mapas de susceptibilidad y peligrosidad para corrientes de derrubios en Andorra....
  • Geotechdata, 2020. https://www.geotechdata.info/parameter/ (accessed 1 February...
  • J.N. Goetz et al.

    Evaluating machine learning and statistical prediction techniques for landslide susceptibility modeling

    Comput. Geosci.

    (2015)
  • F. Guzzetti et al.

    Landslide hazard evaluation: a review of current techniques and their application in a multi-scale study, Central Italy

    Geomorphology.

    (1999)
  • F. Guzzetti et al.

    Rainfall thresholds for the initiation of landslides in central and southern Europe

    Meteorol. Atmos. Phys.

    (2007)
  • F. Guzzetti et al.

    Estimating the quality of landslide susceptibility models

    Geomorphology

    (2006)
  • F. Guzzetti et al.

    Probabilistic landslide hazard assessment at the basin scale

    Geomorphology

    (2005)
  • R.W. Healy et al.

    Using groundwater levels to estimate recharge

    Hydrogeol. J.

    (2002)
  • Cited by (96)

    View all citing articles on Scopus
    View full text