Search

Relationships between rockfall and environmental factors have been the focus of considerable research over the last several decades. Many prior studies have not quantified rockfall volume due to challenges in volume estimation without the use of modern remote sensing methods, and the role of longer-term conditioning factors on rockfall activity is often not considered. As part of a broader effort to evaluate rockfall hazard for slopes along the heavily trafficked Interstate-70 corridor in Colorado, this study takes advantage of a high temporal resolution lidar-derived rockfall database consisting of over five years' worth of rockfall data at a particular cut slope to evaluate the relationship between rockfall activity and environmental factors. The combination of monitoring duration and spatiotemporal monitoring resolution used allows for evaluation of both triggering and conditioning factors. Correlations are evaluated between rockfall metrics and weather metrics at a seasonal scale as well as the scale of individual monitoring intervals. Based on the results, it is interpreted that surficial rock material is weakened each Winter by freezing processes, large amounts of rockfall occur in the Spring as a result of the slope thawing out, elevated rockfall activity continues over the Summer as surficial rocks destabilized by freezing processes in the previous Winter are fully dislodged, and rockfall activity significantly reduces in the Fall due to the number of unstable blocks available to fall being significantly reduced. From comparison with the literature, we conclude this cycle is relatively consistent with rockfall trends observed in alpine environments, while more temperate regions exhibit notably different trends in rockfall seasonality. • Lidar monitoring recorded 1134 rockfalls at the Floyd Hill rock slope. • Winters with high air freezing index have larger rockfalls in the following seasons. • Rockfall is triggered by rain in Summer, mostly independent of rainfall amount. [ABSTRACT FROM AUTHOR]

Many large-scale landslides have occurred along the active Pingding-Huama fault in Zhouqu segment, Gansu, China. To better understand the failure mechanisms of these landslides, we use the Yahuokou landslide as a detailed case study. Field investigation was conducted to retrace the kinematics of the landslide and corresponding timeline of triggering mechanics. Dynamic triaxial tests were conducted to quantify the effect of rainfall and rockfall load on the physical and mechanical character of the landslide materials with in-situ stress level. Numerical simulation was used to evaluate the landslide stability under rainfall and rockfall load. The results show that a smaller initial rockfall of the limestone blocks along the upper headscarp of the landslide triggered a series of larger failure events that propagated through the greater landslide complex. We proposed that continuous rainfall and this rockfall load increased the pore water pressure and significantly reduced the shear strength parameters of the sliding materials. In addition, the rockfall load destroyed the structure of the shallow soil with buried depth <5 m, increasing the pore volume and water absorption capacity, which may cause the water content of the soil to exceed its liquid limit, and finally promoted plastic flow. Stability calculation further showed that rainfall alone was not sufficient to induce the landslide failure, but rather the coupled action of rainfall and rockfall load was needed. The conclusions drawn from this study outline complex failure mechanics of the Yahuokou landslide and may be helpful in understanding the fault-zone landslides widely distributed along the Pingding-Huama fault. • The failure kinematics of upper-part deposits of the Yahuokou landslide is retraced. • The failure of the gravel clay triggers a series of the failure events of the landslide. • The coupled action of rainfall and collapse load induces the 2019 Yahuokou landslide. [ABSTRACT FROM AUTHOR]

A digitally supported rock mass stability analysis is herein presented for the implementation of non-contact close-range remote surveys against landslide hazard in sheltered environments, such as Marine Protected Areas. An airborne photogrammetric survey was carried out at the Lachea Islet (Sicily), which is part of a small volcanic archipelago with a peculiar geological history, hosting unstable rock masses belonging to the first eruptive stages of Mount Etna. The digital model built for this study allowed extracting the discontinuity orientation data, performing a kinematic analysis along the recognized discontinuity clusters, and quantifying the volume of potentially unstable boulders. Laboratory tests allowed achieving the main physical-mechanical parameters of involved rocks, which were used to perform a quantitative stability analysis under both deterministic and probabilistic approaches, based on the extracted digital data. Finally, a 3D rockfall trajectory analysis was carried out by simulating statistical rockfall path on the digital model built for the study. Results highlighted a poor stability in the studied islet sector, which is worsened by taking into account the high seismicity that characterizes the area. The main tourist elements of the islet must be considered at risk, as they are located within the simulated trajectories of the landslides. The methodological protocol carried out for this study, based on conventional procedures but using remotely derived and digitally processed information, can be considered a useful tool in the frame of nature-based solutions for hydro-meteorological hazards connected to climatic changes. • Close-range remote surveys were used to support rock mass stability analysis. • Digital outcrop models are built from UAV photogrammetric surveys. • A Non-invasive surveying protocol is shown as a study tool at a nature reserve. • 3D rockfall simulations allowed evaluating the landslide impact on elements at risk. [ABSTRACT FROM AUTHOR]

The devastating Kahramanmaraş earthquake sequence occurred on February 6, 2023. Two main events, M w 7.8 and M w 7.5 occurred 9 hours apart, affected 11 cities in Turkey, and subjected an area of ∼90,000 km2 to shaking levels known to trigger landslides (peak ground acceleration > 0.08 g). Extensive landsliding was expected given the hilly terrain affected by this significant ground shaking—about 15% of the topography is steeper than 20°—but was not initially apparent in early satellite imagery, mostly because of obscuring snow that fell just after the earthquakes. However, after a more detailed investigation using high-resolution satellite images, aerial photos, and a field survey, we confirmed that this earthquake sequence did, indeed, trigger numerous landslides. In this study, we present those findings and provide a preliminary characterization of the spatial distribution, general characteristics, and dominant types of landslides and hillslope deformation triggered by the earthquake sequence. We mapped 3673 coseismic landslides, mostly concentrated in the northern half of the impacted area. Rock falls are the most abundant landslide type, but bedrock rotational landslides, translational slides and lateral spreads are also numerous. Surface rupture through mountainous terrain caused several large, and in some cases fatal, landslides. Incipient landslides and ground cracks are also widespread, especially in the north. Lithology, spatial variability of ground shaking, and topographic relief appear to be the main variables controlling the spatial distribution of coseismic landslides. There are few detailed studies of earthquake-triggered landslides in arid and semi-arid regions such as this one, nor for such complex earthquake sequences. Therefore, this contribution provides valuable information for future hazard and modeling efforts in arid and semi-arid regions. • The Türkiye-Syria earthquake sequence triggered 3673 landslides. • The ground shaking, topography and lithology control the landslide distribution. • Surface rupture through mountainous terrain caused several large and fatal landslides. [ABSTRACT FROM AUTHOR]

Distributed acoustic sensing (DAS) using fiber-optic cables shows promise for rockfall monitoring by potentially offering high spatial resolution and wide coverage at lower costs. However, further research is needed to fully evaluate DAS performance for rockfall applications. This study assesses DAS capabilities for rockfall monitoring by analyzing DAS and seismometer recordings of 38 artificially triggered rockfalls with block masses ranging from 139.6 to 638 kg. Results demonstrate strong correlations between DAS waveforms and spectra with colocated seismometers, validating DAS data for detecting rockfall impacts. Fiber configuration impacted DAS sensitivity: fully in-contact cables improved signal-to-noise ratios and detection of high frequencies from 55 to 86 Hz. Analyses of seismic waves identified three distinct rock block propagation modes associated with intact, fragmenting, and collapse-triggering movements. While DAS signal feature values agreed with seismometers, no clear mass correlation emerged within the examined range. However, spectral features like frequency centroid (48.8 ± 6.5 Hz) and predominant frequency (41.0 ± 12.1 Hz) exhibited spatial stability across channels, indicating potential for event identification. Preliminary attempts to locate rockfall sources from DAS arrays proved feasible but require optimization. Overall, enhanced fiber coupling, identification of propagation modes, spatially-dependent analyses, and dense-array source location may enable DAS to complement sparse seismic networks for higher resolution, cost-effective rockfall monitoring. These insights could inform the development of improved rockfall monitoring strategies leveraging distributed fiber-optic sensing. • DAS signals match colocated seismometers, validating impact detection. • Fully coupled fiber improves DAS sensitivity/high-frequency detection up to 86 Hz. • Three distinct rock block propagation modes identified, linked to movements. • DAS spectral features spatially stable, indicating event identification potential. • Preliminary dense-array source location using DAS demonstrates feasibility. [ABSTRACT FROM AUTHOR]

The analysis of three-dimensional point cloud data is becoming one of the most used approaches to assess instabilities processes affecting rock slopes. With the increased collection of point cloud data, there is an increasing demand for rapid computational point cloud segmentation techniques to format data for rock fall risk analysis. However, there is a scarcity of semantic segmentation research focusing on geological data and the practical issues they pose. In this study, we propose a novel application to automatically define the spatial extent and intensity of weathered rock slope areas. The studied rock slopes are in Hegra Archaeological Site (al-Hijr / Madā ͐ in Ṣāliḥ), the first UNESCO World Heritage area inscribed in the Kingdom of Saudi Arabia (KSA) in 2008. Here, cavernous weathering processes, with honeycombs and tafoni features, affect the carved rock slope and increase the rockfall hazard. With the spread and improvement of close-range methods, such as Terrestrial Laser Scanner (TLS) and Unmanned Aerial Vehicle Digital Photogrammetry (UAV-DP), it is possible to obtain fast and accurate high-resolution point cloud data of the studied scenario, i.e., rock mass. This paper presents the automated mapping results obtained by a combination of explainable artificial intelligence (XAI) algorithms and three-dimensional geometric analysis of a point cloud using the geometric feature tools available for free on CloudCompare (CC) open-source software. Three-dimensional data were obtained using the UAV-DP technique. The proposed workflow allows, using random forest modelling (RF), to quantify the cavernous weathering extent and estimate the relative intensity on the rock slope, faster than using traditional time-consuming manual segmentation processes. The RF model was able to accurately forecast the cavernous weathering regions (accuracy of up to 85%), with a slight propensity to underpredict, as indicated by the somewhat low recall (∼ 63%), and a limited number of false positives, as indicated by the high precision (up to 84%). This approach is easily reproducible at low cost in terms of computational, human, and financial resources. The results highlight the reliability of the proposed method for the preliminary assessment of cavernous erosional features with the aim of future assessment, monitoring, and deeper analysis of rock-carved slopes. [Display omitted] • UAV-DP surveying provides 3D point cloud of weathered rock slope. • Point cloud 3D geometric features allow to identify weathered areas. • AI-aided cavernous weathering assessment using geometric features. • Automated cavernous weathering detection on UNESCO rock-carved cultural heritage. • Rockfall source areas detection linked to weathering feature. [ABSTRACT FROM AUTHOR]

Defining the relationship between volume and return period is critical when estimating the risk of rockfalls and/or rock avalanche, especially during continued global warming at high altitudes that threatens rock wall stability. Characterizing the volume-frequency relationship based on historical datasets is, however, limited by observation and quantification biases, which have not received enough attention. Here, to monitor recent activities for the Brenva Spur (Mont-Blanc massif, Italy) that is also a rock avalanche scar and estimate the return period of future rock failures based on the volume-frequency relationship (and the corresponding uncertainty), a structure-from-motion photogrammetric survey was conducted from 2017 to 2021. 39 rockfall sources with volumes ranging from 11 to 13,250 m3 were identified within the scar. The total failure volume is 22,438 m3, with an associated erosion rate of 15.5 mm/year, indicating very active morphodynamics possibly linked to the permafrost evolution in the spur. The volumes were characterized by a negative power-law that fits significant two events in 2016 (3.4 × 104 m3) and one in 1997 (2.0 × 106 m3) remarkably well, and the randomness of the fit was evaluated by a Monte Carlo approach. 7 potential failure scenarios ranging from 3.1 × 104 m3 (S 1) to 4.8 × 106 m3 (S 7) were defined according to a structural analysis and the sloping local base level concept. Their extrapolated return periods derived by the power-law fit indicate a longer return period for the maximum failure scenario than for the smaller scenarios. S 1 has a 50% chance of occurring every 3 years, while S 7 has a 50% chance of occurring every 31 years. Though the median return period of S 7 is 31 years, the 95% and 68.2% confidence intervals range from 8 to 399 years and 14 to 93 years, respectively, which reflects a high level of uncertainty but is realistic when considering global warming, progressive rock failure, etc. In addition to characterizing recent rock failure activities in high mountains, this study offers a preliminary examination of the return periods of some extreme scenarios and provides primary data for risk management in mountainous areas that are very sensitive to global warming. • Updated rock failure database by helicopter based SfM photogrammetry survey for Brenva Spur. • Detailed rockfall sources analysis linking geo-structures on the Brenva avalanche scar. • Extreme failure scenarios definition by discontinuity characterization and the SLBL concept. • Failure scenario return period estimation by volume-frequency relationship considering the uncertainties. [ABSTRACT FROM AUTHOR]

A preliminary, yet timely characterization of rockfall hazards is essential for effective emergency response and mitigation measures. Such characterization is, however, particularly challenging for events occurring on high-mountain slopes in remote locations, for which precursory information remains often undetected. Large rockfalls are known to produce seismic signatures that, if properly interpreted, provide key information on the magnitude and destructiveness of the event. To achieve this, we establish a seismic signal-based assessment scheme and demonstrate its capability by taking a large event – the 5 April 2021 Hongya rockfall (Sichuan, China) – as a case study. First, we show how a rockfall can be distinguished from an earthquake and a rockslide by analyzing its seismic signatures. Then, we demonstrate how the kurtosis-based method can be used to rapidly detect the initiation of a rockfall and determine the seismic wave velocity accordingly, as well as how the arrival-time-based location method can be used to locate the event. The rockfall volume can be estimated from the magnitude of radiated seismic energy. Furthermore, we characterize the different phases in seismograms and link them to the typical stages of rockfall, i.e., precursory small rockfalls, impact and fragmentation, sliding and entrainment, and gravity deposition stages. Finally, a discussion and suggestions are provided for further improving the robustness of the assessment scheme. Our results show that the seismic signal-based scheme presented in this study is suitable for characterizing large rockfalls and has the potential for rapid and effective emergency management. • A seismic signal-based assessment scheme for characterizing rockfalls is established. • The scheme is suitable to rapidly identify, detect, locate and characterize rockfalls. • Kurtosis function can be used to classify the dynamic process of a rockfall. [ABSTRACT FROM AUTHOR]

Rockfall tracking and mass estimation are critical for effective hazard response. While seismic techniques have been applied to the field, systematic analysis remains lacking. In this study, we conducted rockfall experiments using seismic networks to investigate tracking and mass estimation issues. Two methods, time-based and amplitude-based source location methods, were used to identify initial rock-ground impact points. The high-fitness areas (>0.95) overlapping between the two methods showed a location error of <8.7 m for initial rock-ground impact points. However, the combined methods could not track bounce points on the slope. To address the limitation, we further defined a seismic energy ratio, a ratio between target and reference stations. When a reference station was far away from rockfall, the rising or decreasing of high-frequency (70–99 Hz) seismic energy ratios reflected the rock approaching or departing from the target station, respectively. Additionally, the peak low-frequency (1–40 Hz) seismic energy ratio of ground stations indicated the timing of the initial rock-ground impact. Successful implementation of these methods depends on optimal station distribution. The study recommends that different source-to-station distances with triangular station geometry and stations at different elevations provide a reliable location determination and straightforward rockfall tracking. Furthermore, the lowest frequency(f L) satisfied the power spectral threshold of −100 dB during the rockfall moving process correlated with the rock mass(M), M = 38,989F L -2.03 with R2 of 0.62. The relation was practical for estimating rockfalls' mass for future monitoring in the same slope. Our experiments demonstrate that this seismic technique can effectively assist in rockfall assessment. • Hybrid time and amplitude-based methods offer reliable results for an initial rock-ground impact point. • A high-frequency seismic energy ratio can tack a rockfall's movement. • A peak of low-frequency seismic energy ratio indicates an initial rock-ground impact point. • The lowest frequency caused by rockfall's bounce or impact is associated with rock mass. [ABSTRACT FROM AUTHOR]

The radon problem in Kinsarvik, Ullensvang municipality in Western Norway, is among the most severe in the world. Annual average indoor radon concentrations as high as 56,000 Bq/m3 have been measured. The problem affects >100 houses, where average yearly indoor radon concentrations of 4340 Bq/m3 has been reported. Indoor radon concentrations vary during the year, which has been explained by thermally induced flows of radon-bearing air through a porous ice-margin deposit. Updated geological knowledge and new mapping methods (LIDAR data) have allowed for a re-examination of the problem. There is significant evidence that does not fit with the established geological understanding that classifies the Kinsarvik deposit to be an ice-marginal moraine. Instead, the landforms and internal structures of the deposit are characteristic of a rock-avalanche deposit. The LIDAR data suggests a rockslide deposit of approximately 50 million m3. Airborne and ground gamma-ray spectroscopy show elevated uranium content in the granitic bedrock north and east of Kinsarvik. These uranium-bearing rocks are also found in an open pit in the Kinsarvik deposit. Terrestrial cosmogenic nuclide dating of four surface boulders shows an average age of 10,900 ± 600 years. This suggests the rock avalanche fell immediately after the last deglaciation of Western Norway. Our results support that the radon problem in Kinsarvik is caused by uranium bearing bedrock and a porous rock-avalanche deposit that emanates radon, which in turn exhibits an alternating, temperature driven flow direction. • Extremely high indoor radon concentrations are documented in Kinsarvik Norway. • Granitic rocks have an increased uranium content that produces radon gas. • The actual deposit is proved to be a result of a rock avalanche. • The rock avalanche has crushed the uranium bearing rocks and makes for radon emanate easy. • The combination of uranium bearing rock and a rock avalanche can be a worldwide problem. [ABSTRACT FROM AUTHOR]

Abstract This study investigates the mechanism of rockfall impact against a granular soil buffering layer above a concrete/rock shed via numerical simulations by discrete element method (DEM). The soil buffering layer is modeled as a loose packing of polydisperse spherical particles, while the bottom concrete/rock shed is simulated by a layer of fixed particles. The rock blocks of various ellipsoidal shapes are represented by an assembly of densely packed and bonded spherical particles. The DEM model was employed to investigate the dynamic interaction between the rock block and the soil buffering layer, including the impact force, penetration depth and bottom force. During the rockfall impact, the force chains occur immediately at the impact area and then propagate radially downward into the soil buffering layer. The force waves vanish gradually as most of the impact energy was absorbed and dissipated by the loose buffering soil particles. The numerical results show that the maximum impact force acting on the rock block increases, while the corresponding penetration depth decreases linearly with the block sphericity. The maximum force acting on the bottom concrete/rock shed is approximately twice the maximum impact force acting on the rock block, showing apparent force amplification of the soil buffering layer. The ratio between these two forces is almost independent of the rock block sphericity. These findings can finally contribute to the design of effective soil buffering layer for concrete/rock sheds. Highlights • A 3D DEM model has been employed to investigate the impacts of rock blocks against a soil buffering layer. • The sphericity index can be used to quantify the overall shape characteristics of rock blocks. • The maximum impact and bottom forces increase linearly with the rock block sphericity. • The penetration depth decreases linearly with the rock block sphericity. [ABSTRACT FROM AUTHOR]

Abstract Geometric information and discontinuity characterization of rock masses are the key aspects of the analysis of the evolution and failure mechanisms of rockfalls. With the advantageous use of terrestrial laser scanning (TLS), the accurate three-dimensional (3D) spatial information of rock slopes can be obtained without contact. By conducting a fuzzy K-means algorithm in this study, the automatic identification of discontinuity sets is achieved, and dominant occurrences of rock mass discontinuities in each local region can be acquired in great detail. This automatic identification method permits the user to visually identify the discontinuity sets and acquire their spacial distribution features, e.g. occurrences, spacings, trace lengths and their geometric compounding relationships. At the same time, based on the shortest distance (SD) algorithm and the surface-to-plane volume calculation algorithm, the distributions, shapes, volumes and scars of the rockfalls can be accurately detected over the monitored time interval. These methods are able to provide adequate investigations and quantitative assessment for the rockfall failure mechanisms and evolutions of the Hongshiyan post failure rock slope after the 2014 Ludian earthquake. The topography of the landslide surface, mechanical properties, scales and the structural features of the discontinuities have a significant effect on the failure mechanisms, distributions and volumes of rockfalls. The main failure mechanisms of the rockfall investigated in this rock slope can be divided into plane failure, wedge failure and toppling failure. The plane and wedge failures mainly occurred in the landslide surface while the toppling failures are mainly observed along the edge of the post failure slope, which is obviously responded by preferred discontinuities. The reverse-dip stratified structure characteristics and the excavation disturbances during the recovering and slope treatment stage induce numerous progressive and continuous failures of rockfalls. The results are beneficial for the design and optimization of rockfall treatments. Highlights • The accurate 3D surface models of the Hongshiyan remnant rock slope have been established using the TLS technology. • Discontinuity sets are visually identified using the automatic discontinuity identification method. • Distributions, shapes, volumes of rockfalls are accurately detected over the monitored time interval. • Failure and evolution mechanisms of rockfalls have been quantitatively analyzed and assessed. [ABSTRACT FROM AUTHOR]

This article first proposes a literature-based criterion for evaluating the capacity of protection embankments in resisting rockfall-induced impacts. The criterion is intended to help stakeholders and public authorities expediently evaluate the strength of existing embankments. It is intentionally kept simple so that it can be applied to any type of embankment and even in those lacking design documentation. The criterion was developed based on available data from real-scale impact experiments conducted worldwide on various types of embankments. It relates the embankment downhill face deformation to the block-incident kinetic energy, differentiating reinforced and non-reinforced embankments. This criterion is then applied to a set of ninety-eight embankments built in France and Switzerland. While the dynamic loading was seldom considered for their design, it appears that >50% of these embankments meet the criterion. This held evaluation, more importantly serves to identify the nearly one-third of this set for which the application of this criterion suggests an excessive downhill deformation, inviting further investigations with respect to their impact strength and performance expectations [ABSTRACT FROM AUTHOR]

The 2008 Wenchuan earthquake in Sichuan, China triggered numerous landslides in the stricken area. The loose landslide materials retained on mountain slopes and in gullies are prone to reactivation and may transform into debris flows in the rainy season. Nine years after the Wenchuan earthquake, debris flows are still active in Gaojiagou Ravine 16.5 km north of the epicentre. On 14 August 2010, 3 July 2011, 13 July 2013, and 6 July 2016, four large-scale debris flows were triggered by heavy storms in Gaojiagou Ravine. The four debris flows blocked the Minjiang River twice and caused severe damage to nearby villages and reconstruction sites. Several questions arise from these repeated debris flows. Did the level of triggering rainfall change in the four debris flow events? How did the initiation mechanism evolve over time? What are the differences in the runout characteristics in the four events? How did the check dams constructed in 2012 function in the subsequent debris flows? In this paper, we evaluated the loose deposit volumes in Gaojiagou Ravine before and after each of the four debris flows and the runout volumes, and analysed the triggering rainfall intensities, initiation mechanisms and runout characteristics of the four debris flows. The rainfall threshold for the debris flows in Gaojiagou Ravine increased over time, and the initiation mechanisms evolved from landslides to channel-bed failure, and subsequently to channel-bank erosion. The mobility of the debris flows decreased from 2010 to 2016 as the initiation positions moved lower and the particle size of the runout materials became coarser. [ABSTRACT FROM AUTHOR]

Across the world, coastal environments of great landscape, recreational and environmental value are coming under increasing pressure. Within such environments, cliffs are particularly characteristic and unique elements, in which processes of instability develop. Their management requires a transdisciplinary approach ensuring protection of the natural condition of the environment, while at the same time allowing for their continued use and enjoyment. In the case of Barinatxe (Basque Coast), the beach has evolved into a system of foredunes, flanked by cliffs with frequent rockfall processes. This research analyzes the effect of coastal dunes as an element of natural protection. To this end, based on a Digital Terrain Model developed by Terrestrial Laser Scanning (TLS) and in situ geological characterization, 3D modeling has been used to analyze rockfall trajectories and evaluate their runout and energy. These models confirm the protective role of coastal dunes, which act as efficient natural barriers against rock blocks that become detached from cliffs. This is particularly important in areas where the use of tools based on Nature-Based Solutions (NBSs) guidelines is recommended. • Beach foredunes prove to be efficient natural barriers against rockfalls from coastal cliffs • 3D and 2D rockfall modeling allows detailed analysis of rockfall trajectories, runout and energy • Dune protection and conservation demonstrates to be an appropriate natural-based strategy to avoid risks in beach areas. [ABSTRACT FROM AUTHOR]

Two destructive rockfalls occurred in central Taiwan, wherein falling blocks severely damaged an essential bridge of a mountain expressway. A series of field investigations were conducted to recognize the trajectories of the rockfalls and gather information on the falling blocks and bridge damage. According to the obtained field data, a hybrid model of the discrete element method (DEM) and finite element method (FEM) was developed to assess the rockfall impact on the bridge. A DEM analysis was first performed to simulate the rockfall process, which agreed with the actual condition. A detailed FEM was developed with strain-rate-dependent material models to investigate the process, which caused severe damage to the bridge and protective structures under the rockfall impact based on the calculated velocity and kinetic energy of rock blocks obtained from the DEM analysis. The simulated damage mechanisms agreed well with the field observations. The simulated maximum impact force was 4.5 times higher than the original design load and caused severe damage to the bridge structure. The thorough and objective investigation results of this rockfall event and the rigorous hybrid DEM-FEM analysis framework can be essential guidelines for authorities to formulate future hazard mitigation strategies for similar infrastructures under the threats of rockfall impacts. • A rockfall event severely damaged an essential bridge in central Taiwan. • This study proposes a hybrid model of DEM and FEM to assess the rockfall impact on the bridge. • The simulated damage mechanisms agree well with the field observations. [ABSTRACT FROM AUTHOR]

• Optimal arrangement of rockfall protection structures combined with risk analysis. • Risk quantification considering several rock block sizes and their frequencies. • Hazard and exposure analysis based on three-dimensional trajectory simulation. • Determination of positions and types of countermeasures that minimize road risk. • Optimization subject to budget constraints and changes in traffic volume. The purpose of this study is to introduce a novel framework that aids in the installation of rockfall protection structures for the risk reduction of rockfall-related damage. In this study, we propose a computational method that performs risk analysis for a road and plans an optimal arrangement of rockfall protection structures. First, a rockfall trajectory simulation is performed using a three-dimensional numerical analysis method to obtain sufficient data for risk analysis. Then, the risk to motorized traffic during a certain period is quantified considering the relationship between rock block volume and rockfall frequency, the movement of rock blocks, the vulnerability of elements at risk, and some other factors. Finally, a knapsack problem that aims to minimize the total risk on a road while satisfying a budget constraint is solved. The proposed method is applied to an actual site and represents that it can derive a layout for protection structures to maximize the cost-benefit performance depending on the road traffic and budget conditions. The obtained results of the structural protection plan can be used to support road managers in decision making. [ABSTRACT FROM AUTHOR]

Rockfall source areas identification of high steep fractured rock slope is a challenging task for hazard mitigation. Considering this difficulty, a new method integrating discontinuity-based threshold slope angle and rockfall trajectory analyses is introduced in this study and applied to a railway bridge in China. First, abundant discontinuity data are collected by means of the modern unmanned aerial vehicle (UAV) digital photogrammetry technique. Then, a threshold slope angle used to delineate rockfall-prone areas is comprehensively determined by probabilistic kinematic and block theory analyses. Next, physically based rockfall trajectory simulation with consideration of mean block size is used to distinguish the hazard level of each release point. Finally, rockfall sources causing moderate and high risks to the train tunnel entrance at the base of high steep fractured rock slope are determined using a semiquantitative risk matrix approach. The final rockfall source map can provide a quantitative reference for hazard mitigation. At the end of this paper, some issues associated with the proposed approach are discussed, such as parameter importance analysis of threshold slope angle, comparison between the two probabilistic approaches and uncertainties in rockfall source areas identification. • A method for rockfall sources identification of jointed rock slope is proposed. • This method integrates threshold slope angle and rockfall trajectory analyses. • Probabilistic approaches are developed to determine a final threshold slope angle. • A multiscanline method is presented to determine the volume of a launched block. • Trajectory dense and sparse areas are used to distinguish different hazard levels. [ABSTRACT FROM AUTHOR]

In this paper, we introduce the Rockfall Activity Index (RAI), a point cloud-derived, high-resolution, morphology-based approach for assessing rockfall hazards. With the RAI methodology, rockfall hazards are evaluated in a two-step procedure. First, morphological indices (local slope and roughness) are used to classify mass wasting processes acting on a rock-slope. These classifications are then used with estimated instability rates to map rockfall activity across an entire slope face. The rockfall hazard is quantified as the estimated annual kinetic energy produced by rockfall along 1-m length segments of a slope face. Field assessment of the RAI method at multiple study sites indicates that the morphology-derived classification and hazard assessment routines provide results that closely match the observed behavior and performance of rock slopes. [ABSTRACT FROM AUTHOR]

The presence of boulders, a result of rockfall events, at the foot of cliffs testifies for a mass wasting process which is widely observed and is often considered to be triggered by earthquakes. Here we investigate boulders under an overhanging cliff with an emphasis on the principal factors that control boulder size. We used a conceptual cantilever beam model to predict the expected maximum length of the overhanging cliff. We then compared these calculated lengths to those of the field observed boulders and the length of the overhanging cliff ledge. Our observations show that the actual lengths of the boulders and overhanging cliff are much shorter than expected from the model. As the study area is located in a tectonically active region, we first suggested solving this discrepancy by adding the possible effect of ground motion on the size of a boulder. Detailed pseudo-static and dynamic analysis of the effect of earthquakes on the failure of rock beams shows that boulder length may not significantly be shortened by induced ground motions. We therefore conclude that strong earthquakes, although triggering rockfalls, are not expected to control the resultant size of the boulders found at the base of overhanging cliffs. We alternatively suggest that the maximum length of a boulder is strongly controlled by the rock strength. We conclude that the maximum boulder size that may fall during an earthquake event can be predicted by calculating the cantilever critical length including the effect of the rock size on its tensile strength. This, together with the measured rock discontinuities in an overhanging cliff suggests an upper-bound for the volume of the falling boulders, a crucial factor in assessing rockfall hazards. [ABSTRACT FROM AUTHOR]

On November 17, 2015 a strong, shallow earthquake (Mw = 6.5) occurred on the island of Lefkada along a N20 ± 5°E, east-dipping strike-slip fault with right-lateral sense of slip. The event triggered environmental effects that were mainly mapped at the western part of the island while moving towards the eastern part, the severity of the earthquake-induced failures decreased. The most characteristic geological effects that were triggered by the earthquake were slope failures including rock falls, rock slides and landslides. The spatial density of the failures follows an exponential increase with slope-angle and decreases for values larger than 55°. Small-size liquefaction phenomena such as sand craters and ejection of fine-grained sand through ground cracks were documented at the coastal area of Vassiliki. The epicentral intensity was assessed based on the definitions of Environment Seismic Intensity scale as VIII ESI-07, and the macroseismic epicenter (VIII up to IX) was located at the coastal area of Egremnoi (west of Athani village, south Lefkada). In addition, comparing the severity of earthquake-induced failures on the main road network, triggered by the 2003 and 2015 events, it is concluded that the behavior of the protection measures such as fences and walls, could be generally characterized as satisfactory. On the contrary, the similarity of the structural failures triggered by the 2003 event with the ones triggered by the 2015 earthquake, documented at the port of Vassiliki, indicates that the post-2003 constructed port facilities failed to guarantee the functionality of the port of Vassiliki. [ABSTRACT FROM AUTHOR]

Montserrat Mountain is located near Barcelona in Catalonia, in the northeast of Spain, and its massif is formed by conglomerate interleaved by siltstone/sandstone with steep slopes very prone to rockfalls. The increasing number of visitors in the monastery area, reaching 2.4 million per year, has highlighted the risk derived from rockfalls for this building area and also for the terrestrial accesses, both roads and the rack railway. A risk mitigation plan has been launched, and its first phase during 2014–2016 has been focused largely on testing several monitoring techniques for their later implementation. The results of the pilot tests, performed as a development from previous sparse experiences and data, are presented together with the first insights obtained. These tests combine four monitoring techniques under different conditions of continuity in space and time domains, which are: displacement monitoring with Ground-based Synthetic Aperture Radar and characterization at slope scale, with an extremely non-uniform atmospheric phase screen due to the stepped topography and atmosphere stratification; Terrestrial Laser Scanner surveys quantifying the frequency of small or even previously unnoticed rockfalls, and monitoring rock block centimetre scale displacements; the monitoring of rock joints implemented through a wireless sensor network with an ad hoc design of ZigBee loggers developed by ICGC; and, finally, monitoring singular rock needles with Total Station. [ABSTRACT FROM AUTHOR]

A gravelly scree slope in the Meretschibach catchment, a location in the Swiss Alps in the vicinity of Agarn, canton Valais, has been observed to deform downslope at up to 0.5 m p.a. The potential instabilities at this site include surficial landslides, some of them originally thought to be triggered by an increase in pore water pressure with a subsequent loss of shear strength as a consequence of rainfall infiltration and rockfalls. A programme consisting of monitoring, laboratory testing and investigation was developed, to perform a thorough soil characterisation needed in order to produce a realistic ground model. The long-term geotechnical monitoring included in situ soil temperature, suction as well as volumetric water content measurements using dielectric permittivity and time domain reflectometry (TDR) sensors. This data was complemented by electrical resistivity tomography (ERT) to provide extensive knowledge on the depth to bedrock and to validate the volumetric water contents in specific locations. The datasets are completed by recordings from two nearby weather stations. Seasonal changes of precipitation and temperature were reflected in corresponding trends in all measurements. A comparison of volumetric water content records was obtained using capacitance and time domain reflectometry (TDR) sensors with ERT, yielding reasonable agreement. The resulting ground model, which integrates all currently available parameters, delivers the essential information and boundary conditions for predicting and validating slope instabilities in the future, using numerical and physical modelling. [ABSTRACT FROM AUTHOR]

Scenario-based risk assessment for rockfalls, requires assumptions for different scenarios of magnitude (volume). The magnitude of such instabilities is related to the properties of the jointed rock mass, with the characteristics of the existing unfavourably dipping joint sets playing a major role. The critical factors for the determination of the maximum credible rockfall volume in a study site, the Forat Negre in Andorra, are investigated. The results from two previous analyses for the rockfall size distribution at this site are discussed. The first analysis provides the observed size distribution of the rockfall scars, and it is an empirical evidence of past rockfalls. The second one, calculates the kinematically detachable rock masses, indicating hypothetical rockfalls that might occur in the future. The later gives a maximum rockfall volume, which is one order of magnitude higher, because the persistence of the basal planes is overestimated. The tension cracks and lateral planes interrupt systematically the basal planes, exerting a control over their persistence, and restricting the formation of extensive planes and large rockfall failures. Nonetheless, the formation of basal planes across more than one spacings of tension cracks is possible and small step-path failures have been observed too. Concluding, the key factor for the determination of the maximum credible volume at the study-site is the maximum realistic length of the basal planes, penetrating into the rock mass, their spacing, and, if applied, the contribution of the rock bridges to the overall rock mass resistance. [ABSTRACT FROM AUTHOR]

• A 3D discrete elements model (DEM) is used to perform rockfall trajectory analysis. • Sphero-polyhedra enable the use of realistic shapes for blocks and topography. • Analysis rely on real-scale experimental data, such as passing velocities and end points. • An indicative shape coefficient is proposed to coarsely estimate the areas where blocks pass through and stop. • Simulation and experimental deposition zones are given in terms of lateral dispersion and run-out distances. Trajectory analysis is often needed to handle rockfall hazards. It is crucial to understand and to refine ground and rock interactions during a rockfall, which can be related to the elements involved. They include the topology, ground and rock nature, but also the volume and shape of the falling blocks. This work focuses on replicating field observations using a 3D discrete elements model (DEM) in order to further analyze the possible rockfall deposition areas, which may not be accessible due to the limited number of experimental data available. The numerical model implements blocks of more realistic shapes that were reconstructed from in situ blocks obtained by photogrammetry. The dissipation of kinetic energy at the collision point is suitably managed. In the experimental campaign, dozens of boulders (rock block) releases were conducted on two slope profiles of a quarry located in Authume (France). Block passing heights, velocities and runout distances were assessed at specific ground points. We analyzed lateral spreads, propagation distances, and energy balances computed for 3 different block geometries. These numerical results were confronted with experimental observations. Although time-consuming compared to lumped mass and rigid body dynamics models where the impact duration is zero, the DEM used in this work is versatile thanks to an explicit consideration of geometrical effects throughout the life of multiple contacts. It allows to simulate quite accurately a multitude of configurations. Among the properties of the terrain and the blocks, geometric features are shown to be crucial. For the sake of efficiency, a simplified shape coefficient based on block's elongation is proposed. [ABSTRACT FROM AUTHOR]

Rockfall risk management has traditionally been built upon scenarios based on statistical analysis of historic frequency-magnitude rockfall patterns which allows for an estimation of the return period of rockfalls within given ranges of magnitude. Τhe availability of rockfall databases is thereby of crucial importance to quantify rockfall activity. In recent years, engineering geologists have been able to leverage light detection and ranging (LiDAR) technology with automated change detection and clustering workflows, for spatially accurate rockfall monitoring and inventorization. This wealth of digital data generates new research prospects for the development of an artificial intelligence (AI) system able to analyze change detection sequences and map spatio-temporal rockfall patterns by precisely capturing the rock slope evolution in near-real-time. However, prerequisite to such an advancement is the need to efficiently automate the process of classifying surface changes corresponding to rockfall. This task typically requires intensive human effort and is prone to multiple sources of human error like subjectivity, expertise, and experience that may differ from person to person involved in a particular project. This paper investigates the potential of integrating sophisticated deep learning architectures into dynamic rockfall database population processes, with the goal of relieving experts from the daunting task of manually classifying rockfall events within large loads of change detection data. Deep neural networks based on the pioneering PointNet and PointNet++ architectures for 3D point cloud learning are developed for this purpose based on a 5-year change detection database consisting of >8000 rock slope clusters of identified change for training, with scanning intervals ranging from 5 days to 6 months. The models are tested on the 536 clusters from the two last data acquisitions to simulate the real monitoring situation and subsequently on the most frequent of the campaigns to increase the probability of working with single-event clusters. The best-performing model achieves an accuracy of about 89% and 84% on the last and shortest campaign, respectively. The optimized deep learning models are further evaluated on a geologically different rockfall database achieving almost 93% accuracy in a location where discrete geomorphologic features such as steep rock outcrops and erosion channels are present. The study shows that although it is challenging to achieve generalization in rockfall detection, site-specific training of the proposed deep learning architecture can lead to high-level performance and support further advancements in rockfall risk management. • LiDAR monitoring provides detailed 3D point cloud models of rock slope deformations. • High-level feature learning from point clouds is achieved with modern deep learning. • Rockfall detection can be approached as a binary point cloud classification problem. • Global shape feature learning is suggested compared to local texture features. • The prediction confidence increases with the cluster volume. [ABSTRACT FROM AUTHOR]

Rockfall causes a large number of accidents and fatalities in steep environments. A realistic quantification of rockfall risk is thus crucial for an effective prevention of damages and loss of lives. The estimation of rockfall and block volumes for different return periods thereby remains a major challenge. In this paper, we present a straightforward rockfall frequency model (RFM: Rockfall Frequency Model) and its application at 8 different sites at 7 locations in the Swiss Alps. The RFM assumes that the magnitude-frequency relationships of rockfall events and blocks follow a power law. The parameters of this distribution are estimated based on a simple classification of rock structures and on field inventories. Beside the block volume frequency, which is very sensitive to the consideration of large rockfall events, the frequency of rockfalls with at least one block with a minimum volume, is determined. The block size distributions measured in this study were well captured by power laws. The rockfall and block volumes calculated with the RFM were generally slightly higher than the scenarios of the official hazard assessments. The uncertainty analysis, however, revealed a high variability of the release scenarios with respect to the parameters of the RFM, increasing with the return period. Both, the rockfall volumes and the block volumes, are particularly sensitive to the estimated exponent of the power law distribution of the rockfall events. Nevertheless, the proposed RFM provides an objective and transparent approach to derive magnitude-frequency relationships of rockfall events and individual blocks even if historical inventories are missing or insufficient and is thus a promising alternative to merely expert-based approaches. • A straightforward rockfall frequency model allows for an objective estimation of rockfall and block release frequencies. • Block size distributions measured in the field were well fitted by power laws. • The uncertainty in the calculated release scenarios increases with the return period. [ABSTRACT FROM AUTHOR]

In the mountainous areas passed by the Sichuan–Tibet Railway, rockfall is one of the important geological hazards with characteristics of unpredictable and sudden occurrences. For high and steep slopes, this study employs nap-of-the-object photogrammetry with the characteristic of close flight distance to reconstruct a 3D geological model, which is clear and qualified for discontinuity identification. For the rockfall investigation of areas where unstable blocks are widely distributed, this study proposes a new methodology for 3D rockfall risk statistical analysis to conduct well-directed rockfall simulations and perform a rockfall analysis on a tunnel portal. This methodology first narrows down the rockfall source regions through a field investigation and by slope aspect. Then, the critical source areas that threaten the tunnel portal are determined by preliminary simulation results. Finally, unstable blocks within the critical source areas are interpreted, and well-directed simulations are performed. The rockfall risk analysis on the tunnel portal is conducted on the basis of trajectory distributions and kinematic parameters. This methodology greatly simplifies the rockfall simulation process and reduces the workload of unstable block identification and interpretation. As a consequence, in the Kangyuqu area, several unstable blocks are identified for simulations. Results indicate that rockfalls reaching the tunnel portal have a kinetic energy range of 300–20,000 kJ, a velocity range of 8–30 m/s, and a maximum jump height of 25 m. • Nap-of-the-object photogrammetry is employed to reconstruct a 3D geological model. • Feasibilities of discontinuity interpretation and unstable block detection are analyzed. • A new methodology combining statistics for rockfall risk analysis is proposed. • Detailed rockfall analysis on a tunnel portal of Sichuan-Tibet Railway is performed. [ABSTRACT FROM AUTHOR]

In this work, we present the results of a rockfall trajectory study performed on the south-western slope of Mt. Catiello (Sorrento Peninsula, southern Italy). Such a study develops within a multi-methodological approach which integrates different types of remote sensing data and techniques. Specifically, ground-truth data (e.g., rock mass geo-structural information, rock block inventory) were generated by geologically-supervised interpretations of high-resolution virtual outcrop models (VOMs). These data were then used for reconstructing the in-situ fractured rock mass attributes of the Mt. Catiello peak, as provided by a Discrete Fracture Network (DFN) model, and to prepare the subsequent numerical simulations of rockfall trajectories. The resulting rockfall scenarios are consistent with the ground-truth data, both in terms of size and spatial distribution. Thus, we believe that the proposed approach can be effectively applied to other areas, characterized by similar geological features but higher levels of exposure and vulnerability. • A multi-methodological approach for rockfall simulations: from slope face to valley • Integration of UAV-SfM and satellite-based data for rock mass and debris analysis • Obtaining ground-truth information from different types of virtual outcrop models [ABSTRACT FROM AUTHOR]

Although drapery systems are commonly used to control rock fall hazard on steep slopes, some elements of their design such as the interface friction between the mesh and ground surface are still difficult to quantify in practice. This technical note presents a new test procedure designed to study the mechanism of rock–mesh interaction in the laboratory. A series of tilt tests and tests with increasing loads were performed to study the effects that number and type of rock–mesh contacts, slope angle, and mesh characteristics had on mesh–rock interaction. The obtained data indicated that the process of interlocking between the mesh and rocks could increase the mesh's resistance to failure as well as decrease the force acting on the anchors during accumulation of rock debris or snow. Some changes to the current guidelines regarding the selection of interface friction angle are also suggested. [ABSTRACT FROM AUTHOR]

The analysis of the failure processes and mechanisms of rockfalls that are associated with underground mining activities is presented. The study area is located in Kaiyang Phosphorite Mine in Southwest China, where the geological condition is dominated by anti-dip slopes with layers inclined backward into the natural slope with decreasing strength in the rock mass from the upper (dolomite-rich) to the lower (shale-rich) strata. The analysis is based on field investigation and laboratory experimental study using a gravitational simulation device. The results support the proposed failure mechanism by demonstrating the process of failure from the deformation of the roof and floor in the stopes, to the development of surface cracks near the top of slopes, and eventually to the formation of rockfalls as the surface cracks propagate along pre-existing joints. The rockfalls are classified into one of three failure modes: crack–toppling, crack–sliding, and crack–slumping, in which the failures are governed by the corresponding characteristics of the rock mass structure. The study of the failure process and their spatial and temporal correlation with the underground workings reveals warning signs or indicators of impending slope instability. Improved understanding of the failure process and indicators can aid in early identification and timely warning of geohazards in phosphorite mines. [ABSTRACT FROM AUTHOR]

A storm can trigger numerous shallow landslides on hilly terrains. This paper presents a distributed cell model for presenting regional shallow landslide movement on a three-dimensional digital terrain. The model consists of five components; namely, a digital terrain module, a spatial rainfall distribution module, an infiltration analysis module, a slope stability evaluation module, and a movement prediction module. The locations and volumes of landslides are analyzed first through rainfall infiltration and slope stability analyses. The detached material is assumed to move along the steepest path from one cell to a lower cell. Empirical equations that are developed based on local landslide inventories are adopted as a landslide movement cessation criterion. The effect of landslide size on travel distance is considered by grouping unstable cells that are bounded at least at one side or corner. The uncertainties in soil shear strength parameters and the landslide movement process are considered in determining the possible bounds of travel distance. The movement analysis method is applied to a 164.5 km 2 hilly terrain in the Wenchuan earthquake zone to test its performance in presenting regional shallow landslide movement. The method predicts the volume of landslides reasonably well. The observed and computed movement traces and deposition locations of the detached material also agree reasonably well. The volume of individual landslide and travel distance may be substantially underestimated if the size effect is not considered. [ABSTRACT FROM AUTHOR]

A number of rainfall threshold models for triggering landslides have been developed in either empirical or physical manner. The objective of an empirical rainfall threshold model is to establish a relationship between rainfall intensity and duration, which is a line on a log–log diagram, although the mechanism of the relationship following the power law is not yet well understood. The sufficiency of rainfall and landslide data is also a significant concern for empirical models. A simplified physically based rainfall threshold model is developed by coupling the Mohr–Coulomb law and Darcy's law based on the assumptions that the slope has an infinite length and that discrete water diffusion is physically based. The power-law mechanism for the relationship between rainfall intensity and duration for the initiation of a landslide is investigated using this physically based model. The rainfall threshold calculated with the model shows that the threshold follows the power law for short rainfall durations whereas it gradually tends towards a constant value for long rainfall durations. Case studies have shown that a plausible threshold for regions that lack sufficient landslide and rainfall data can be obtained via the model. The model is also applied to cases which occurred in regions geologically typical of Fujian Province, southeast China. The landslide distribution predicted by our model accords with the interpretation of high resolution satellite imagery, and this suggests that the proposed physically based model also effectively determines rainfall thresholds for triggering regional landslides. [ABSTRACT FROM AUTHOR]

The paper presents a novel optimal plane arrangement of stabilizing piles in terms of the provided half simplified flattened ellipsoid model, which can be used to describe the three-dimensional characteristics of sliding mass for colluvial landslides. By studying the friction soil arching effect between the adjacent stabilizing piles, a reasonable pile spacing model for stabilizing piles was deduced in consideration of the driving force and shear strength of sliding mass as well as the dimension of pile cross-section. The results indicate that the net pile spacing is in direct proportion to the height of pile cross-section and shear strength parameters of sliding mass, but in inverse proportion to the driving force intensity. Consequently, the net pile spacing should steadily increase from the major slip profile to the profiles near the boundaries accordingly, which can be called the non-uniformly spaced arrangement principle of stabilizing piles for colluvial landslides. In view of the scale factor and groundwater in the different cross sections of landslide, the improved formula of stability coefficient based on the Fellennius method was proposed accordingly. Furthermore, the concept of stability limit was put forward to confine the rational arrangement region for stabilizing piles; consequently, the region beyond the rational arrangement region is not necessary to set piles any more. The case study of the Erliban landslide in the Three Gorges Reservoir Region of China shows that the quantitative optimal plane arrangement scheme based on the proposed reasonable pile spacing model and stability limit only requires 25 stabilizing piles rather than 31 stabilizing piles in the conventional uniformly spaced arrangement scheme, with an obtainable savings of 19.4% in the number of stabilizing piles. [ABSTRACT FROM AUTHOR]

In the Thompson and Fraser River valleys, rockfall and rockslides are a hazard for railway traffic in the region. In June 2013, a 2600 m 3 rockslide occurred in the White Canyon, causing damage to infrastructure and interruption to service. In the year preceding this failure, the slope was monitored using terrestrial Light Detection And Ranging (LiDAR) as part of the Railway Ground Hazards Research Program initiative. The purpose of this study is to evaluate a hazard management strategy for transportation corridors exposed to steep slopes involving the identification of rockfall precursors prior to significant slope failures. For this failure, deformation was observed in the seven months preceding collapse with an accelerating period occurring in the final 28 days. A distinct pattern of smaller rockfalls was observed around the perimeter of the deforming rock block as well as loss of surficial material over the slides tension crack. This study shows that with LiDAR monitoring, precursors to rock slope failures can be identified allowing for better rock slope hazard management by focusing attention on areas with the highest probability of failure. [ABSTRACT FROM AUTHOR]

We deployed seismic and infrasound sensors at a historically active cliff in Yosemite Valley for the purpose of detecting and locating rock falls at local (< 1 km) distances and demonstrate the potential for using these techniques for real-time rock fall monitoring. The project ran for two winters: the first deployment was a system feasibility study consisting of a single station with a geophone and an accelerometer; the second deployment was a network of seven stations at four different locations with the addition of infrasound sensors. We demonstrated that small (< 20 m 3 ) rock falls are detectable at distances of several hundred meters, individual impacts can be identified, and seismic waves are generated prior to the first main impact for some rock falls. We also found that infrasound is viable and compliments seismic, especially for locating events. We correlated the data with environmental conditions and extracted information about the initiation, triggering, and dynamics of rock falls. A major part of the research effort was the development of a triggering algorithm and criteria for distinguishing rock falls from thousands of seismic triggers. Twelve rock falls were identified in the continuous seismic recording by searching for triggers and comparing them with known rock falls and other forms of seismic activity. Physical evidence or reports of rock falls exist for only eight of the twelve rock falls that we identified; thus, we have demonstrated that instrumented monitoring can significantly augment the detection of rock falls even in heavily-trafficked areas such as Yosemite Valley. Six of the rock falls appear to be related to each other as an ongoing instability, while the rest appear to be independently occurring events. After we identified the individual rock fall events, we focused on characterizing the seismic data in terms of timing, frequency, and P/S/Rayleigh wave phases in order to develop a set of characteristic parameters indicative of rock fall signals. [ABSTRACT FROM AUTHOR]

This paper presents a procedure for tracking rockfall trajectories and extracting kinematic parameters from both the impacts and the resultant fragments. A set of full scale rockfall experiments was performed in a quarry located in Vallirana, Barcelona (Spain). The study site was chosen due to the presence of a rigid discontinuity surface, inclined at 42° in the middle of the slope, whose configuration was expected to favor the breakage of the blocks. The trajectories of the blocks released and of the resultant fragments were recorded with three video cameras. A C++ program was specifically developed to track the 3D trajectory of blocks and fragments, and measure velocities before and after the impact. Two different modules were implemented, one for the blocks that break and one for those that do not. The trajectory of a non-fragmented block is obtained by comparing it to its 3D model. In this way, both the center of mass position and the orientation of the block are tracked. For fragmented blocks, the local coordinates of the fragments determined from the images are converted to terrain coordinates using the program we developed. A total of 16 blocks and 36 rock fragments after impact were tracked. The parameters obtained were georeferenced and linked to a common system of 3D terrestrial coordinates. The captured parameters allow obtaining the velocity distribution of fragments, the coefficient of restitution, and energy balance for the blocks that break. To our knowledge, this is the first attempt to capture kinematic parameters of rock fragments that result from the impact and breakage of rock blocks in full-scale tests. Although the analysis of the rockfall fragmentation phenomenon is beyond of this work, we have compared the performance of the fragmented and unbroken blocks. To this purpose, we have built 3D models of the rock fragments generated using images captured with a drone. The results indicate that blocks that fragment show higher rebound velocities and coefficients of restitution than the blocks that do not although there exists a certain overlap between the two groups. Despite the experiment is carried out on the same discontinuity surface and with small variations in the impact velocities, impact kinetic energies and impact angles, the coefficients of restitution obtained present a wide range of values, both for the blocks that break and for those who do not. The number of tested blocks is too small to draw generalizable conclusions, but they highlight the stochastic nature of the rebound process and the necessity to consider additional parameters for its understanding. Finally, the results confirm the relation between the dissipated energy and, especially the impact energy and the new area created by fragmentation. Furthermore, the blocks that hit the ground with the face are those that generate the most new area while those that hit the vertex generate less. • Characterizing block fragmentation. • Capturing block's impact-rebound characteristics. • Stereo photogrammetric analysis (3D) of block's trajectory. • Real-scale rockfall testing. • Obtaining block and fragments masses, velocities, energy balance and restitution coefficients. [ABSTRACT FROM AUTHOR]

Inventory of pre-earthquake and earthquake triggered landslides is used to provide insight into the interplay between climatic and tectonic forcing in the development of the rock slopes of the Skolis Mountain, in the NW Peloponnese. Aerial photograph analysis and surface mapping indicate that the Skolis Mountain is characterized by long-term climatically and tectonically controlled rock falls forming taluses. Temporally these taluses show a slow progressive inflation in surface area from 1945 to 2007. However, the post-earthquake surface area of the rock falls increased three times. Similarly 75 rock fall sites before the earthquake, increased into 89 after the Movri Mountain earthquake (Mw 6.4). In addition, during the earthquake a series of isolated rock falls descended Skolis slopes causing threat of the Santomerion village and blocking significant part of the dirt roads around it. These boulders are clustered in three areas beyond the base of taluses. The rock slope failures are controlled by a complex array of discontinuities that are conveniently related to rock mass classification following the geological strength index. These discontinuities are associated with joints and faults caused during the formation of the Hellenides fold- and thrust-belt, and/or related tectonic damage. We infer that a dense pattern of fractures in limestone plays a crucial role in the reactivation of movement within the rock falls during the 2008 Movri Mountain earthquake. All these data are used to define two borders, the taluses base and the rock fall hazard border beyond the base of taluses. For defining these borders we use the angle β drawn from the boulders' release zone down its maximal runout points. Our results indicate that the border defined by the β = 33° corresponds to the climatically driven rock falls while the β = 24° border is defined as the boulders' maximum runout during earthquakes. [ABSTRACT FROM AUTHOR]

On August 31st, 2013, high intensity rainfall caused a severe rockfall disaster in Badouzih, Keelung. This study combines the results of detailed field investigation and process-based as well as empirical-based method to point out unsafe road sections. The mechanisms and the characteristics of motions of rockfall are analyzed in the field. Two intersecting joints and the bedding plane divided bedrock into blocks. Weathering and the infiltration of rainfall accelerated the erosion process and precipitated ruptures. When the rainfall reached its highest intensity which is 94.5 mm/h in this event, a great deal of water on the slope surface eroded the matrix, causing instability and leading to the failure of the rock blocks. The process of rockfall took 23 s and can be divisible into 4 stages including roll, fall, bounce, and roll. Process-based simulations are adopted in this case and several critical road sections to understand the characteristic of movements. With the empirical-based energy line angle analysis, potential road sections which may affect by rockfall are showed and their energy are also estimated for the purpose of mitigating rockfall impact, reducing similar disasters in this area. [ABSTRACT FROM AUTHOR]

We present a general framework for probabilistic landslide hazard analysis. With respect to other quantitative hazard assessment approaches, this probabilistic landslide hazard analysis has the advantage to provide hazard curves and maps, and to be applicable to all typologies of landslides, if necessary accounting for both their onset and transit probability. The method quantifies, for a given slope location, the exceedance probability of being affected by a landslide with a specific local intensity within a reference time interval, i.e. the hazard curve, under the common assumption that landslides behave as a Poisson process. Hazard maps are calculated, reducing the hazard curve to single values by choosing a fixed probability of exceedance following standards or regulation requirements. The method is based on the assessment of a landslide onset frequency, a runout frequency for long-runout landslides, and the local definition of landslide intensity, which can be expressed through different parameters, according to landslide typology. For long runout landslides, the runout and spatially-varying intensity and uncertainty are considered. Hazard curves and maps play a fundamental role in the design and dimensioning of mitigation structures, in land planning and in the definition of risk and hazard management policies. Starting from the general framework, we apply the methodology for rockfall hazard analysis, and we test it in an area affected by the Christchurch 2011 earthquake, New Zealand, which triggered a large number of rockfalls, killing five people. [ABSTRACT FROM AUTHOR]

Abstract: This paper presents a numerical approach for the design of rockfall catchfence based on the study of the rockfall stabilization works at Route 5 state highway in NJ, US where severe rockfalls have been periodically occurring. For the optimal design of rockfall catchfence, statistical rockfall analysis has been performed to estimate critical impact conditions and impact energy in the study area. Then, the design criterion of the rockfall catchfence has been validated by means of 3D finite element analysis featuring explicit dynamic capabilities. A series of parametric studies indicate that the rockfall barrier system can be optimally designed by controlling the design parameters such as the stiffness and configurations of net barrier system and impact conditions. The paper illustrates how detailed numerical analyses can be applied practically to optimize the design of the rockfall barrier system under a given field condition. [Copyright &y& Elsevier]

Supervised Machine Learning (ML) can be used to automatically interpret remote sensing data in engineering geology, with applications for rockfall and landslide characterization. However, supervised algorithms typically require very large training databases from which to learn predictive relationships, and there is little guidance on how to construct such a database in an earth science context with high temporal and spatial heterogeneity. This study builds a supervised classifier to perform basic rock slope characterization on Structure from Motion (SfM) point clouds collected under a variety of conditions. Eight datasets were collected in Colorado and Utah, USA, with multiple different sensor platforms (terrestrial and aerial), rock types, seasons, and lighting conditions, and each dataset was manually labeled to identify regions of vegetation, rock, soil, talus, and snow. A total of 2560 Random Forest models were built with different combinations of training datasets and combinations of geometric and color features. Most models were able to identify vegetation and rock with high accuracy (median F scores of 86% and 68% respectively), but performance for soil, talus, and snow was overall much poorer, and the median overall accuracy of generalized classifiers was 60%. Many characteristics of the training data were found to have significant effects on generalization accuracy, indicating that training datasets must be curated to be applicable to specific data collection parameters, seasons, lighting conditions, and geological settings. We conclude that high accuracy generalized results can be obtained, but the ML model must be carefully constructed, and its limitations acknowledged. • Photogrammetry datasets from 8 slopes were used to train Random Forest classifiers. • Accuracy was highly influenced by the combination and number of sites used in training. • Aggregating multiple sites for training generally improves performance. • Sites with high point density and similar geology and lighting to the target are best. • High accuracy is possible, but empirical testing and domain knowledge are required. [ABSTRACT FROM AUTHOR]

Rockfalls pose a substantial threat to ground transportation, due to their rapidity, destructive potential and high probability of occurrence on steep topographies, often found along roads and railway routes. Approaches for the assessment of rockfall susceptibility range from purely phenomenological methods and statistical methods, suitable for modeling large areas, to purely deterministic ones, usually easier to use in local analyses. A common requirement is the need to locate potential detachment points, often found uphill on cliffs, and the subsequent assessment of the runout areas of rockfalls stemming from such points. Here, we apply a physically based model for the calculation of rockfall trajectories along the whole Italian railway network, within a corridor of total length of about 17,000 km and varying width. We propose a data-driven method for the location of rockfall source points based on expert mapping of potential source areas on sample representative locations. Using empirical distributions of gridded slope angle values in source areas mapped by experts, we derived probabilistic maps of rockfall sources in the proximity of the railway network, regardless of a particular trigger. Source areas act as starting points of simulated trajectories within the three-dimensional model STONE. The program provides a pixel-by-pixel trajectory count, covering 24,500 km2 and representing the largest homogeneous application of the model to date. We classified the raster map into a vector susceptibility map, analyzing the railway track as a collection of segments, for which we provide segment-wise rockfall susceptibility. Eventually, we considered an equivalent graph representation of the network, which helps classifying the segments both on the basis of rockfall susceptibility and the role of each segment in the network, resulting in a network-ranked susceptibility. Both maps are useful for subsequent hazard assessment, and to prioritize safety improvements along the railway, at national scale. • We performed physically based rockfall susceptibility analysis along the Italian railway, over an area of size 25,400 km2 • Sample rockfall sources mapped by experts helped devising a probabilistic identification of potential sources • Locations of sources was a key input for three–dimensional simulation of rockfall runout areas within the STONE model • We classified of railway segments in rockfall susceptibility classes, and with a new joint index including network properties [ABSTRACT FROM AUTHOR]

Rock instabilities can represent major risks for local populations depending on their geographical location. Even if different mechanisms are known to trigger rockfalls such as precipitation, seismic activity and freezing, many rockfalls occur during periods when such trigger mechanisms are absent. Some recent studies have pointed out the role of thermal cycling in crack initiation or propagation. Here we use data from nearly five years of field monitoring of a French limestone cliff. Evolution of the aperture of joints were followed as well as displacements and temperature deep in the rock mass. We show that seasonal thermal variation affects the rock at depths of up to 6 m and subcritical cracking can probably appear even at this depth. • we use data from nearly 5 years of field monitoring of a perched cave in a limestone cliff • 4 fissurometers were positioned to quantify the aperture of the joints, to evaluate the relative movement of the blocks • 2 borehole extensometers with 3 measurements points each inform of the global behaviour of the rock mass • temperatures were measured in a cave (surface temperature) but also at 2 m and 6 m depth • seasonal thermal variation affects the rock up to 6 m deep; subcritical cracking can probably appear even at this depth [ABSTRACT FROM AUTHOR]

In the process of impact-rebound, rockfalls with an initial rotational speed (RS) often undergo a violent transformation between rotational energy and translational energy, which may affect the coefficient of restitution (COR). However, owing to the complexity of experimental simulations and the influence mechanism, this issue is often neglected during research. Therefore, in this study, four representative blocks (ellipsoid, cylinder, cone with a spherical base, and disc) were selected, and impact tests of each block rotating around different axes with various RSs were conducted using a specially developed device. Based on the impact dynamics theory, the RS has no significant effect on the tangential COR (R t), while it may exhibit a positive correlation with the R t when the size of the block size is large enough. When the block rotates around the X-axis (approximately normal impact), there is no obvious relationship between the RS and the normal COR (R n). Conversely, when the block rotates around the Y-axis (often non-normal impact), the RS influences the R n , but it is difficult to determine in what manner and to what extent. Thus, to clarify the influence mechanism of RS on the R n for non-normal impacts, a comprehensive variable, the impact posture coefficient (IPC), which comprehensively considers the RS, block shape, and impact posture, was introduced, and the impact-rebound characteristics of the blocks were classified and discussed. Under the condition of bouncing after only one impact, the IPC and R n values exhibited a positive linear correlation when the block mass centre (C) is behind the contact point (CP), whereas the opposite behaviour was observed when the C is in front of the CP. Meanwhile, under the condition of bouncing after two impacts, the R n and IPC values had a positive linear correlation. These findings explain the influence mechanism of RS on the COR and provide a new idea for the accurate prediction of the impact-rebound processes of rotating irregular rockfalls. • Effect of rotational speed (RS) on coefficient of tangential restitution is not significant only for small size block. • For non-normal impact, RS does affect coefficient of normal restitution (R n) but it's hard to directly affect Rn alone. • Introducing impact posture coefficient to quantify the effect of RS on R n for non-normal impact. • Study on block's impact- rebound characteristics by classified discussion method. • Clarify the influence mechanism of RS on coefficient of restitution. [ABSTRACT FROM AUTHOR]

Abstract: Structures used in rockfall protection are designed on the basis of the expected impact energy. This quantity is usually assessed using commercial lumped mass models that stochastically predict possible block trajectories on a given slope. In New South Wales, Australia, it is estimated that rockfall hazard involves values of energy which are much lower than those in Europe, Canada or the USA. However, this view has not been supported by any systematic study across the whole state. Such a study is presented in this paper. It applies to the five geological situations that are the most prone to rockfall in eastern Australia. Previous experimental findings, relating to block size distribution or restitution coefficients (reviewed herein), have been used to perform this statistical analysis. A newly formulated lumped mass program, which incorporates a relationship between the normal restitution coefficient and the impact angle, allows the adoption of normal restitution coefficients in excess of unity for impact angles lower than 30°. The results confirm that in three out of the five geological situations, the 95th percentile of impact energy is lower than 200kJ, and less than 2000kJ for the other two situations. [Copyright &y& Elsevier]

Abstract: Catastrophic failure of large land masses, which generate landslides, rockfalls and debris avalanches, can have hazardous consequences extending far beyond the source. Observations show that the mobility of such events depends strongly on the volume for volumes larger than 106 m3, with many different processes invoked to explain higher mobilities (hypermobility) for both terrestrial and extraterrestrial events. Although the mobility of large events has been extensively studied, there is no generally accepted mechanism for predicting extreme travel distances because the underlying physical processes are poorly understood. Here we show using physical and rheological arguments that the wide scatter observed for very large mass wasting events in all environments collapses to a single relationship between event volume or inundation area and mobility. Hypermobility is defined to be the reciprocal of the effective friction coefficient μ e , where the scale-dependent μ e is derived analytically as a function of the mechanical, volumetric and topographical parameters of the flow. The dominant term in the coefficient is the degree of fluidization involved in the flow; our results show that fluidization is limited in extraterrestrial events, that significant fluidization occurs in non-volcanic and volcanic events, and fluidization dominates submarine events. This analysis demonstrates that fluidization is associated with long run-out distances, and that the degree of fluidization can be predicted by the volume, and physical and topographic parameters. The methodology is simple, physically-based and validated with datasets of very large terrestrial and extraterrestrial avalanche events. We demonstrate that the effective Coulomb friction rheology and the hypermobility function are applicable to avalanche events of any size, providing an opportunity to simulate past and/or potential huge landslide and debris avalanche events, run-out distances, destructive impact and assessment of risk. The model can be used to estimate the overrun area and volume in terms of known mobility data. [Copyright &y& Elsevier]

Abstract: The analysis of the three-dimensional rockfall scar geometry provides clues for the understanding of the failure mechanisms acting on cliffs, of the conditioning factors, and on the frequency of the events. In this paper, a supervised step-by-step methodology is presented for establishing the statistical magnitude–frequency relation of rockfall scar volumes, using a point cloud from Terrestrial Laser Scanner (TLS) data. The methodology includes a procedure for identifying discontinuity surfaces, calculating the areas of those which are exposed, and the height of rockfall scars. In the estimation of the rockfall scar volume a key issue is the consideration of the minimum spacing of the discontinuity sets to differentiate between step-path surfaces and undulated ones. Having obtained the distributions of both the basal area and height of the scar across the slope, the volume of the rockfall scars is calculated stochastically by multiplication of these two parameters by means of a Monte Carlo simulation. Both distributions of the basal area and of the rockfall scar volume are found to be power-law, with the exponent b ranging from 0.9 to 1.2. The relation obtained might be used as a first approach of rockfall magnitude–frequency curves in large cliffs. [Copyright &y& Elsevier]

Abstract: This paper describes the geomechanical back-analysis of the flexural failure of an overhanging limestone slab (volume=68 m3) that collapsed on 26/01/1999 on the abandoned Cellina Valley state road SS 251 (western Friuli, NE Italy). The survey carried out on site examining the contact surfaces between the block and the slope has ascertained that the rock slab was connected to the stable rock mass by means of a single great rock bridge. The detailed reconstruction of the slab geometry and of the actual restraint conditions acting before the rockfall permitted an estimation of the stress state by using a 3D finite element code (Strand7). The finite element numerical model, referred to the restrained cross-section measured on site (A res =2.82 m2), gives a maximum value of σ t =5.19 MPa for the tensile stress, which is in good agreement with the estimated value for the characteristic strength mobilized at rupture by the intact rock material (T 0 =3.5–5.5 MPa). Considerably lower maximum tensile stress (σ t =2.25 MPa), smaller than one half, is obtained if the whole contact surface is assumed as resisting surface (A res =10.50 m2) and the rock bridge is neglected. [Copyright &y& Elsevier]

Abstract: A modified version of the Colorado Rockfall Hazard Rating System (RHRS) was created by adding several geological and climatic factors that have been recently shown to contribute to rockfall. The modified RHRS was then used on slopes distributed throughout the western, mountainous terrain in Colorado, including sites where only cut slopes were rated and sites where both the cut portion and the higher, uncut portion were both judged to contribute to the rockfall hazard. In total, 355 slopes were rated: for crystalline rocks, 147 cut slopes and 77 total slopes, for sedimentary rocks, 43 cut slopes and 31 total slopes, and for block-in-matrix materials, 45 cut slopes and 12 total slopes. Univariate least squares regression was used to find parameters that had a statistically significant influence on the total hazard score. This subset was then analyzed using ordinal logistic regression to rank these parameters against one another to determine which had the most significant influence on the total hazard score. Finally, statistically significant parameters from these regressions were analyzed using stepwise regression to obtain predictive equations to estimate the total hazard score for each of the slope types analyzed based on the scores of a few parameters. The stepwise regression produced equations in which the total hazard scores can be estimated from four of the RHRS parameters, reduced from the original 12 or 18 parameters, depending on slope type, with R 2 values ranging from 68 to 82%. For some cases, equations with more terms and correspondingly higher R 2 values could be applied. Analysis of the equations identified the most important and least important parameters for each type of slope. For all slope types, the presence of launching features and the slope aspect were important, while slope angle, annual freeze/thaw rating, slope water, and friction along discontinuities were not important. While not as detailed as the full RHRS, this approach allows for a more rapid preliminary rating of rock slopes for screening-level evaluation and it helps identify the most basic geologic factors influencing rockfall potential. [Copyright &y& Elsevier]