Background: This paper reviews the classical and some particular factors contributing to earthquake-triggered landslide activity. This analysis should help predict more accurately landslide event sizes, both in terms of potential numbers and affected area. It also highlights that some occurrences, especially those very far from the hypocentre/activated fault, cannot be predicted by state-of-the-art methods. Particular attention will be paid to the effects of deep focal earthquakes in Central Asia and to other extremely distant landslide activations in other regions of the world (e.g. Saguenay earthquake 1988, Canada). Results: The classification of seismically induced landslides and the related 'event sizes' is based on five main factors: 'Intensity', 'Fault factor', 'Topographic energy', 'Climatic background conditions', 'Lithological factor'. Most of these data were extracted from papers, but topographic inputs were checked by analyzing the affected region in Google Earth. The combination and relative weight of the factors was tested through comparison with well documented events and complemented by our studies of earthquake-triggered landslides in Central Asia. The highest relative weight (6) was attributed to the 'Fault factor'; the other factors all received a smaller relative weight (2-4). The high weight of the 'Fault factor' (based on the location in/outside the mountain range, the fault type and length) is strongly constrained by the importance of the Wenchuan earthquake that, for example, triggered far more landslides in 2008 than the Nepal earthquake in 2015: the main difference is that the fault activated by the Wenchuan earthquake created an extensive surface rupture within the Longmenshan Range marked by a very high topographic energy while the one activated by the Nepal earthquake ruptured the surface in the frontal part of the Himalayas where the slopes are less steep and high. Finally, the calibrated factor combination was applied to almost 100 other earthquake events for which some landslide information was available. This comparison revealed the ability of the classification to provide a reasonable estimate of the number of triggered landslides and of the size of the affected area. According to this prediction, the most severe earthquake-triggered landslide event of the last one hundred years would actually be the Wenchuan earthquake in 2008 followed by the 1950 Assam earthquake in India -considering that the dominating role of the Wenchuan earthquake data (including the availability of a complete landslide inventory) for the weighting of the factors strongly influences and may even bias this result. The strongest landslide impacts on human life in recent history were caused by the HaiyuanGansu earthquake in 1920 -ranked as third most severe event according to our classification: its size is due to a combination of high shaking intensity, an important 'Fault factor' and the extreme susceptibility of the regional loess cover to slope failure, while the surface morphology of the affected...
Summary Origins of ancient rockslides in seismic regions can be controversial and must not necessarily be seismic. Certain slope morphologies hint at a possible co-seismic development, though further analyses are required to better comprehend their failure history, such as modelling the slope in its pre-failure state and failure development in static and dynamic conditions. To this effect, a geophysical characterisation of the landslide body is crucial to estimate the possible failure history of the slope. The Balta rockslide analysed in this paper is located in the seismic region of Vrancea-Buzau, Romanian Carpathian Mountains, and presents a deep detachment scarp as well as a massive body of landslide deposits. We applied several geophysical techniques on the landslide body, as well as on the mountain crest above the detachment scarp, in order to characterise the fractured rock material as well as the dimension of failure. Electrical resistivity measurements revealed a possible trend of increasing fragmentation of rockslide material towards the valley bottom, accompanied by increasing soil moisture. Several seismic refraction surveys were performed on the deposits and analysed in form of P-wave refraction tomographies as well as surface waves, allowing to quantify elastic parameters of rock. In addition, a seismic array was installed close to the detachment scarp to analyse the surface wave dispersion properties from seismic ambient noise; the latter was analysed together with a co-located active surface wave analysis survey. Single-station ambient noise measurements completed all over the slope and deposits were used to further reveal impedance contrasts of the fragmented material over in-situ rock, representing an important parameter to estimate the depth of the shearing horizon at several locations of the study area. The combined methods allowed the detection of a profound contrast of 70-90 m, supposedly associated with the maximum landslide material thickness. The entirety of geophysical results was used as basis to build up a geomodel of the rockslide, allowing to estimate the geometry and volume of the failed mass, i.e., approximately 28.5-33.5 million m3.
Considering the critical importance of the quality of input data for landslide susceptibility, we investigate the performance improvements that can be achieved by different globally available digital elevation models (DEMs) using different state-of-the-art statistical and machine-learning models. For this purpose we compare the predictive performances achieved using terrain attributes derived from TanDEM-X DEM (12 m resolution and resampled to 30 m), ASTER DEM (30 m), SRTM DEM (30 m), and a DEM (25 m) interpolated from contour lines (1:25.000 map scale), exploiting the capabilities of logistic regression, generalized additive models, random forests and support vector machines. The study was conducted in the Buz au Sector of the Curvature Subcarpathians of Romania, a region highly susceptible to landslides. While the performances varied little among modelling techniques, the use of different DEMs strongly influenced the cross-validation accuracy of landslide susceptibility models. TanDEM-X (12 m) based susceptibility models outperformed models based on the other DEMs (median Area Under the Receiver Operating Characteristics Curve (AUROC) values 0.708-0.730). Models using ASTER-derived terrain attributes showed the poorest predictive capabilities (median AUROC 0.568-0.595). We conclude that the quality of DEMs is of critical importance in landslide susceptibility modelling, and greater efforts should be made to obtain suitable DEM products.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.