Regional-scale modelling of shallow landslides with different initiation mechanisms

Sliding versus liquefaction

José J. Lizárraga, Paolo Frattini, Giovanni B. Crosta, Giuseppe Buscarnera*

*Corresponding author for this work

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Rainfall-induced landslides can result from different soil failure mechanisms each leading to slope instabilities characterized by varying times and depths of occurrence. This paper discusses the performance of a physically-based model for the spatially-distributed analysis of landslide susceptibility. The model allows the simultaneous assessment of two failure modes: frictional slips and liquefaction-induced flowslides. The underlying mechanical formulation relies on stability principles for unsaturated soils, while infiltration is treated as a transient process described by the Richards equation. The model has been applied to a series of landslides occurred in 1998 in Campania (Italy), using laboratory data to constrain the input parameters. The analyses provide spatial and temporal patterns of landslide initiation consistent with field evidences, in that they capture most of the affected landslide source areas and predict a temporal evolution of instabilities in agreement with reported failure times. Similar to available regional inventories, the computed fraction of flowslides corresponds to 75% of the total unstable area, while the resulting ratio between success and error indices is twice as large as those reported in other studies for the same site. Parametric analyses conducted by varying the hydraulic conductivity (Ks) have pointed out a strong interplay between the dynamics of infiltration, the mode of slope instability and the depth of failure. In particular, it has been found that low values of Ks promote shallow flowslides due to moisture perching and high saturation at the surface, while larger values of Ks favor deeper infiltration fronts and frictional slips initiating at higher suction. The comparison between these scenarios showed that simulations based on calibrated hydraulic properties encompass the widest distribution of failure depths, thus reproducing more accurately the typical landslides of the site.

Original languageEnglish (US)
Pages (from-to)346-356
Number of pages11
JournalEngineering Geology
Volume228
DOIs
StatePublished - Oct 13 2017

Fingerprint

Landslides
Liquefaction
liquefaction
sliding
landslide
Infiltration
modeling
infiltration
Soils
Richards equation
Hydraulic conductivity
failure mechanism
hydraulic property
temporal evolution
suction
Failure modes
Rain
hydraulic conductivity
Moisture
soil

Keywords

  • Infiltration
  • Liquefaction
  • Pyroclastic soils
  • Shallow landslides
  • Unsaturated soils

ASJC Scopus subject areas

  • Geotechnical Engineering and Engineering Geology
  • Geology

Cite this

@article{28032b4802884023b47c5adeb0c1a5ef,
title = "Regional-scale modelling of shallow landslides with different initiation mechanisms: Sliding versus liquefaction",
abstract = "Rainfall-induced landslides can result from different soil failure mechanisms each leading to slope instabilities characterized by varying times and depths of occurrence. This paper discusses the performance of a physically-based model for the spatially-distributed analysis of landslide susceptibility. The model allows the simultaneous assessment of two failure modes: frictional slips and liquefaction-induced flowslides. The underlying mechanical formulation relies on stability principles for unsaturated soils, while infiltration is treated as a transient process described by the Richards equation. The model has been applied to a series of landslides occurred in 1998 in Campania (Italy), using laboratory data to constrain the input parameters. The analyses provide spatial and temporal patterns of landslide initiation consistent with field evidences, in that they capture most of the affected landslide source areas and predict a temporal evolution of instabilities in agreement with reported failure times. Similar to available regional inventories, the computed fraction of flowslides corresponds to 75{\%} of the total unstable area, while the resulting ratio between success and error indices is twice as large as those reported in other studies for the same site. Parametric analyses conducted by varying the hydraulic conductivity (Ks) have pointed out a strong interplay between the dynamics of infiltration, the mode of slope instability and the depth of failure. In particular, it has been found that low values of Ks promote shallow flowslides due to moisture perching and high saturation at the surface, while larger values of Ks favor deeper infiltration fronts and frictional slips initiating at higher suction. The comparison between these scenarios showed that simulations based on calibrated hydraulic properties encompass the widest distribution of failure depths, thus reproducing more accurately the typical landslides of the site.",
keywords = "Infiltration, Liquefaction, Pyroclastic soils, Shallow landslides, Unsaturated soils",
author = "Liz{\'a}rraga, {Jos{\'e} J.} and Paolo Frattini and Crosta, {Giovanni B.} and Giuseppe Buscarnera",
year = "2017",
month = "10",
day = "13",
doi = "10.1016/j.enggeo.2017.08.023",
language = "English (US)",
volume = "228",
pages = "346--356",
journal = "Engineering Geology",
issn = "0013-7952",
publisher = "Elsevier",

}

Regional-scale modelling of shallow landslides with different initiation mechanisms : Sliding versus liquefaction. / Lizárraga, José J.; Frattini, Paolo; Crosta, Giovanni B.; Buscarnera, Giuseppe.

In: Engineering Geology, Vol. 228, 13.10.2017, p. 346-356.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Regional-scale modelling of shallow landslides with different initiation mechanisms

T2 - Sliding versus liquefaction

AU - Lizárraga, José J.

AU - Frattini, Paolo

AU - Crosta, Giovanni B.

AU - Buscarnera, Giuseppe

PY - 2017/10/13

Y1 - 2017/10/13

N2 - Rainfall-induced landslides can result from different soil failure mechanisms each leading to slope instabilities characterized by varying times and depths of occurrence. This paper discusses the performance of a physically-based model for the spatially-distributed analysis of landslide susceptibility. The model allows the simultaneous assessment of two failure modes: frictional slips and liquefaction-induced flowslides. The underlying mechanical formulation relies on stability principles for unsaturated soils, while infiltration is treated as a transient process described by the Richards equation. The model has been applied to a series of landslides occurred in 1998 in Campania (Italy), using laboratory data to constrain the input parameters. The analyses provide spatial and temporal patterns of landslide initiation consistent with field evidences, in that they capture most of the affected landslide source areas and predict a temporal evolution of instabilities in agreement with reported failure times. Similar to available regional inventories, the computed fraction of flowslides corresponds to 75% of the total unstable area, while the resulting ratio between success and error indices is twice as large as those reported in other studies for the same site. Parametric analyses conducted by varying the hydraulic conductivity (Ks) have pointed out a strong interplay between the dynamics of infiltration, the mode of slope instability and the depth of failure. In particular, it has been found that low values of Ks promote shallow flowslides due to moisture perching and high saturation at the surface, while larger values of Ks favor deeper infiltration fronts and frictional slips initiating at higher suction. The comparison between these scenarios showed that simulations based on calibrated hydraulic properties encompass the widest distribution of failure depths, thus reproducing more accurately the typical landslides of the site.

AB - Rainfall-induced landslides can result from different soil failure mechanisms each leading to slope instabilities characterized by varying times and depths of occurrence. This paper discusses the performance of a physically-based model for the spatially-distributed analysis of landslide susceptibility. The model allows the simultaneous assessment of two failure modes: frictional slips and liquefaction-induced flowslides. The underlying mechanical formulation relies on stability principles for unsaturated soils, while infiltration is treated as a transient process described by the Richards equation. The model has been applied to a series of landslides occurred in 1998 in Campania (Italy), using laboratory data to constrain the input parameters. The analyses provide spatial and temporal patterns of landslide initiation consistent with field evidences, in that they capture most of the affected landslide source areas and predict a temporal evolution of instabilities in agreement with reported failure times. Similar to available regional inventories, the computed fraction of flowslides corresponds to 75% of the total unstable area, while the resulting ratio between success and error indices is twice as large as those reported in other studies for the same site. Parametric analyses conducted by varying the hydraulic conductivity (Ks) have pointed out a strong interplay between the dynamics of infiltration, the mode of slope instability and the depth of failure. In particular, it has been found that low values of Ks promote shallow flowslides due to moisture perching and high saturation at the surface, while larger values of Ks favor deeper infiltration fronts and frictional slips initiating at higher suction. The comparison between these scenarios showed that simulations based on calibrated hydraulic properties encompass the widest distribution of failure depths, thus reproducing more accurately the typical landslides of the site.

KW - Infiltration

KW - Liquefaction

KW - Pyroclastic soils

KW - Shallow landslides

KW - Unsaturated soils

UR - http://www.scopus.com/inward/record.url?scp=85031104419&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85031104419&partnerID=8YFLogxK

U2 - 10.1016/j.enggeo.2017.08.023

DO - 10.1016/j.enggeo.2017.08.023

M3 - Article

VL - 228

SP - 346

EP - 356

JO - Engineering Geology

JF - Engineering Geology

SN - 0013-7952

ER -