Search

Uplifting mountains are denudated by gravitational deformation, landslides, and glacial or fluvial erosion. We investigated the response of valley slopes to glacial and fluvial erosion along the Kali Gandaki (KG) River, one of the largest rivers that flows through the High Himalaya southward from Tibet to India and diagonally cuts the strikes of gneiss and slate foliations. Satellite images with 2.5 m resolution and 5-m Digital Elevation Models (DEMs) were used for the investigation. Within a 20-km reach of the KG River, we found a series of 11 landslides, which vary from a total area of 0.58 to 12 million m2 each. The foliations generally strike NW–SE and dip at 30° to 50° to the NE, so the left bank slopes are anaclinal and the right bank slopes are cataclinal. On the cataclinal slopes, buckling of beds occurs, and gradual deformation proceeds in slate and two-mica gneiss, while in the more competent calc gneiss, buckling deformation develops into a rock avalanche. Such rock avalanches are represented by the largest Dhampu-Chhoya landslide, which once blocked the KG River and drastically changed the river morphology. Anaclinal slopes of two-mica gneiss were toppled and transformed into rock avalanches on three slopes. The KG River has a prominent fluvial knickpoint, which is currently at an elevation of 1800 m, and it has slope breaks on the left bank that are 200–400 m above the riverbed downstream of the knickpoint. Toppling was observed on slopes higher than slope breaks, which suggests that toppling began before the formation of slope breaks by fluvial incision and that the lower slopes did not adapt to the younger fluvial incision. In contrast, slope breaks are scarcely found on the right bank, probably because slow-moving DGSDs, which followed the river incision, erased them. • The Himalayan mountains respond to erosion based on geological structures. • Buckling and gradual deformation proceed on dip slopes of incompetent rock. • Buckling of competent rock develops into catastrophic failure. • On anaclinal slopes, toppling and catastrophic failure occur. • Upstream migration of a fluvial knickpoint induced a series of mass movements. • Glacial erosion and deglaciation destabilized the feet of dip slopes, inducing buckling deformation. [ABSTRACT FROM AUTHOR]

Abstract: Around 11.5∗106 m3 of rock detached from the eastern slope of the Santa Cruz valley (San Juan province, Argentina) in the first fortnight of January 2005. The rockslide–debris avalanche blocked the course, resulting in the development of a lake with maximum length of around 3.5km. The increase in the inflow rate from 47,000–74,000m3/d between April and October to 304,000m3/d between late October and the first fortnight of November, accelerated the growing rate of the lake. On 12 November 2005 the dam failed, releasing 24.6∗106 m3 of water. The resulting outburst flood caused damages mainly on infrastructure, and affected the facilities of a hydropower dam which was under construction 250km downstream from the source area. In this work we describe causes and consequences of the natural dam formation and failure, and we dynamically model the 2005 rockslide–debris avalanche with DAN3D. Additionally, as a volume ~24∗106 m3of rocks still remain unstable in the slope, we use the results of the back analysis to forecast the formation of a future natural dam. We analyzed two potential scenarios: a partial slope failure of 6.5∗106 m3 and a worst case where all the unstable volume remaining in the slope fails. The spreading of those potential events shows that a new blockage of the Santa Cruz River is likely to occur. According to their modeled morphometry and the contributing watershed upstream the blockage area, as the one of 2005, the dams would also be unstable. This study shows the importance of back and forward analysis that can be carried out to obtain critical information for land use planning, hazards mitigation, and emergency management. [Copyright &y& Elsevier]

Abstract: Landslides are ubiquitous along the Oregon coast. Many are large, deep slides in sedimentary rock and are dormant or active only during the rainy season. Morphology, observed movement rates, and total movement suggest that many are at least several hundreds of years old. The offshore Cascadia subduction zone produces great earthquakes every 300–500years that generate tsunami that inundate the coast within minutes. Many slides and slide-prone areas underlie tsunami evacuation and emergency response routes. We evaluated the likelihood of existing and future large rockslides being triggered by pore-water pressure increase or earthquake-induced ground motion using field observations and modeling of three typical slides. Monitoring for 2–9years indicated that the rockslides reactivate when pore pressures exceed readily identifiable levels. Measurements of total movement and observed movement rates suggest that two of the rockslides are 296–336years old (the third could not be dated). The most recent great Cascadia earthquake was M 9.0 and occurred during January 1700, while regional climatological conditions have been stable for at least the past 600years. Hence, the estimated ages of the slides support earthquake ground motion as their triggering mechanism. Limit-equilibrium slope-stability modeling suggests that increased pore-water pressures could not trigger formation of the observed slides, even when accompanied by progressive strength loss. Modeling suggests that ground accelerations comparable to those recorded at geologically similar sites during the M 9.0, 11 March 2011 Japan Trench subduction-zone earthquake would trigger formation of the rockslides. Displacement modeling following the Newmark approach suggests that the rockslides would move only centimeters upon coseismic formation; however, coseismic reactivation of existing rockslides would involve meters of displacement. Our findings provide better understanding of the dynamic coastal bluff environment and hazards from future subduction-zone earthquakes. [Copyright &y& Elsevier]

Abstract: We review the role of landslides in current concepts of the topographic development of mountain ranges. We find that many studies in this field address basin- or orogen-scale competition between rock uplift and fluvial bedrock erosion. Hillslopes in general, and bedrock landslides in particular, are often assumed to respond rapidly to incision and development of the fluvial drainage network. This leads to a one-sided view of the geomorphic coupling between hillslopes and rivers that emphasizes the fluvial control of hillslopes, but ignores the alternative view that landslides can affect the fluvial network. There is growing evidence that landslides are a dominant source of sediment in mountain belts and that they exert a direct geomorphic control on fluvial processes. Landslides can influence the river network in a variety of ways, from determining basin area and drainage divide positions, to setting streamwise variations in sediment load and calibre. The geomorphic legacy of large landslides on hillslope and channel morphologies may persist for up to 104 yr, adding considerable variability to fluvial erosion and sedimentation patterns over these timescales. We identify a number of questions for future research and conclude that a better understanding and quantification of the geomorphic feedbacks between landslides and river channels builds an important link between short-term (<101 yr) process studies and long-term (>105 yr) landscape evolution models. [Copyright &y& Elsevier]

Abstract: We present sedimentologic and petrographic field evidence of an extremely large (~2.5×109 m3) Late Quaternary rockslide around Dzongri, 24 km S of Kanchenjunga (8585 m a.s.l.). Widespread occurrence of shattered leucogranites and migmatites deposited discordantly on top of undisturbed augen gneisses, and separated by micro-brecciated sliding surfaces indicate rapid movement during a large rockslide event. Occurrence of frictionite (= hyalomylonite) in mineral samples indicates short-term temperatures of ~600 °C in the moving rockslide mass along a 12° sliding plane. From a SW direction of movement we infer a source 8 km NE of the deposit, ~5 km S of Pandim (6691 m a.s.l.), where today a thick leucogranite intrusion marks a zone of weakness and failure of a postulated former mountain crest. The mechanism of movement, the dip of the sliding surface, and the deposit preservation along ridge tops require that the Dzongri rockslide ran out over an ice-filled Prek Chu valley. Hence, we prefer earthquake shaking over undercutting and glacial debuttressing as a possible trigger mechanism. A regional review of other similarly large and catastrophic Himalayan rockslides indicates that leucogranite intrusions are highly conducive to large-scale failure in the Higher Himalayan Crystalline (HHC). Importantly, high erosion rates have rendered this large rockslide deposit undetectable by remote sensing or digital elevation data, which are commonly used for identifying evidence of such extreme events. Quantifying the contribution of such giant landslides in destroying some of the highest relief on Earth will thus need further research attention. [Copyright &y& Elsevier]

In rapidly uplifting regions, deep seated landslides are a prominent natural hazard that influence landscape evolution and are an important source of sediment. They pose a threat to infrastructure, and their failure can deliver sediment that overwhelms or blocks river channels. Consequently, a better understanding of why and where deep-seated landslides occur would help us reduce hillslope erosion, improve water quality, and allow us to better understand how landscapes evolve. Although the slope threshold model provides a framework for understanding how these landslides form, it has several shortcomings. Specifically, it fails to predict the spatial organization of landslides within a region, and it does not explain why landslide-prone portions of a region can exhibit relatively low slopes. To explore these issues, we have used a landslide susceptibility analysis within the Whanganui Basin of the North Island of New Zealand to explore the role that multiple susceptibility factors play in the occurrence of both translational and rotational deep-seated rock slides. Our findings contribute to a growing view of the relative importance of fluvial incision and the more subordinate role that earthquakes play in the development of large, deep-seated landslides. Additionally, our analyses identify the potential influence that widespread deforestation may have had on promoting instability, despite the deep-seated nature of the landslides studied. Most importantly however, we find that where slopes and weak geological structures align, a universal threshold slope does not exist, and many landslides develop on hillsides with below-average slope angles. This suggests that the presence of structural controls on landslide occurrence create a more complex landscape, where hillsides structurally predisposed to landslide occurrence have a much lower threshold slope angle. • We implement a deep-seated landslide susceptibility model within New Zealand. • Seismic factors were not predictive of deep-seated landslide occurrence. • Slope and bedding alignment was predictive of deep-seated landslide occurrence. • Slope and bedding alignment resulted in deviations from a universal slope threshold model. • Reducing river incision may be an effective way to prevent future failures. [ABSTRACT FROM AUTHOR]