Seismic triggering of landslides, Part A: Field evidence from the Northern Tien Shan

Abstract

Abstract. Landslides triggered by strong earthquakes often caused most of the global damage and most of all casualties related to the events, such as shown by the M = 7.7 Peru earthquake in 1970, by the M = 7.6 El Salvador earthquake in 2001 or by the M = 7.4 Khait (Tajikistan) earthquake in 1949. The obvious impact of a landslide on the population is directly related to its movement. Yet, prediction of future failure potential and hence future risk to population is necessary in order to avoid further catastrophes and involves the analyses of the origin of seismic instability. The seismic landslide potential is mainly determined by the interaction between the regional seismic hazard and local geological conditions. At a local scale, seismic factors interfering with geological conditions can produce site-specific ground motions. The influence of such Site Effects on instability is the principal topic of this paper, which is divided into two parts, A and B. The present Part A is concerned with the correlation of field data with observed instability phenomena. Field data were obtained on mainly three landslide sites in the Northern Tien Shan Mountains in Kyrgyzstan, Central Asia. Geophysical prospecting, earthquake recordings, geological observation, trenching and geotechnical tests were the main investigation tools. The collected information gives an insight in the geological background of the slope failure and allows us to roughly infer failure mechanisms from field evidence. A detailed analysis of the susceptibility of a mechanism to specific geological conditions will be shown in Part B.

Seismic triggering of landslides. Part B: Simulation of dynamic failure processes

“…2a). Results of the geophysical and seismological surveys are shown in Part A (Havenith et al, 2003). In particular, the latter revealed strong site effects all around the Ananevo rockslide with amplification values higher than 8 between 0.7 and 2 Hz at the stations close to the mountain crest.…”

“…They further take into account ground ruptures along the slope existing before the earthquake (a more detailed description is given in Part A: Havenith et al, 2003). On the other hand, owing to the weak contrast between the block-material and the fracture shear strength (the latter is only 10% lower than the first, see Table 2), the medium corresponds to a quasi-continuum.…”

“…The data recorded at these sites were described in detail recently (Part A: Havenith et al, 2003). After a brief review of these field observations, this paper (Part B) presents the results of a numerical analysis applied to two of the three investigated slope failures: the Ananevo rockslide and the Suusamyr debris slide.…”

“…On the other hand, it seems that the location of the landslide scarps may provide indications about the origin of the mass movement. Examples cited in Part A (Havenith et al, 2003) Correspondence to: H.-B. Havenith (HB.Havenith@ulg.ac.be) reveal a morphological dependency of the seismic slope failure occurrence.…”

“…As a consequence, in order to examine the influence of various geological and seismic factors on slope movements, it is often necessary to produce numerical models. This paper describes the results of such models as applied to two case studies in Kyrgyzstan: the Ananevo rockslide, located in granite, and the Suusamyr debris slump-flow, situated within soft sediments (see Part A: Havenith et al, 2003). Discrete element modelling (UDEC), adapted both to the discontinuous character of fractured rock and to the heterogeneity of layered mediums, was used.…”