Mapping landslides using automated methods is a challenging task, which is still largely done using human efforts. Today, the availability of high-resolution EO data products is increasing ...exponentially, and one of the targets is to exploit this data source for the rapid generation of landslide inventory. Conventional methods like pixel-based and object-based machine learning strategies have been studied extensively in the last decade. In addition, recent advances in CNN (convolutional neural network), a type of deep-learning method, has been widely successful in extracting information from images and have outperformed other conventional learning methods. In the last few years, there have been only a few attempts to adapt CNN for landslide mapping. In this study, we introduce a modified U-Net model for semantic segmentation of landslides at a regional scale from EO data using ResNet34 blocks for feature extraction. We also compare this with conventional pixel-based and object-based methods. The experiment was done in Douglas County, a study area selected in the south of Portland in Oregon, USA, and landslide inventory extracted from SLIDO (Statewide Landslide Information Database of Oregon) was considered as the ground truth. Landslide mapping is an imbalanced learning problem with very limited availability of training data. Our network was trained on a combination of focal Tversky loss and cross-entropy loss functions using augmented image tiles sampled from a selected training area. The deep-learning method was observed to have a better performance than the conventional methods with an MCC (Matthews correlation coefficient) score of 0.495 and a POD (probability of detection) rate of 0.72 .
Rapid mapping of event landslides is crucial to identify the areas affected by damages as well as for effective disaster response. Traditionally, such maps are generated with visual interpretation of ...remote sensing imagery (manned/unmanned airborne systems or spaceborne sensors) and/or using pixel-based and object-based methods exploiting data-intensive machine learning algorithms. Recent works have explored the use of convolutional neural networks (CNN), a deep learning algorithm, for mapping landslides from remote sensing data. These methods follow a standard supervised learning workflow that involves training a model using a landslide inventory covering a relatively small area. The trained model is then used to predict landslides in the surrounding regions. Here, we propose a new strategy, i.e., a progressive CNN training relying on combined inventories to build a generalized model that can be applied directly to a new, unexplored area. We first prove the effectiveness of CNNs by training and validating on event landslides inventories in four regions after earthquakes and/or extreme meteorological events. Next, we use the trained CNNs to map landslides triggered by new events spread across different geographic regions. We found that CNNs trained on a combination of inventories have a better generalization performance, with a bias towards high precision and low recall scores. In our tests, the combined training model achieved the highest (Matthews correlation coefficient) MCC score of 0.69 when mapping landslides in new unseen regions. The mapping was done on images from different optical sensors, resampled to a spatial resolution of 6 m, 10 m, and 30 m. Despite a slightly reduced performance, the main advantage of combined training is to overcome the requirement of a local inventory for training a new deep learning model. This implementation can facilitate automated pipelines providing fast response for the generation of landslide maps in the post-disaster phase. In this study, the study areas were selected from seismically active zones with a high hydrological hazard distribution and vegetation coverage. Hence, future works should also include regions from less vegetated geographic locations.
Catastrophic landslides characterized by runaway slope failures remain difficult to predict. Here, we develop a physics‐based framework to prospectively assess slope failure potential. Our method ...builds upon the physics of extreme events in natural systems: the extremes so‐called “dragon‐kings” (e.g., slope tertiary creeps prior to failure) exhibit statistically different properties than other smaller‐sized events (e.g., slope secondary creeps). We develop statistical tools to detect the emergence of dragon‐kings during landslide evolution, with the secondary‐to‐tertiary creep transition quantitatively captured. We construct a phase diagram characterizing the detectability of dragon‐kings against “black‐swans” and informing on whether the slope evolves toward a catastrophic or slow landslide. We test our method on synthetic and real data sets, demonstrating how it might have been used to forecast three representative historical landslides. Our method can in principle considerably reduce the number of false alarms and identify with high confidence the presence of true hazards of catastrophic landslides.
Plain Language Summary
Catastrophic slope failures that pose great threats to life and property remain difficult to predict due to the strong variability of slope behavior. As a result, only a limited number of large rock slope failures have been so far successfully forecasted with associated risks mitigated. Here, we propose a novel predictive framework to prospectively and quantitatively detect slope failure precursors with high confidence. Our research sheds light on one of the most challenging questions in landslide prediction: Would an active landslide slowly move or catastrophically fail in the future? Our method adds a new conceptual framework and operational methodology with a significant potential to support existing early warning systems and hence reduce landslide risks.
Key Points
Tertiary creeps of catastrophic landslides accommodate dragon‐kings showing statistically different properties than secondary slope creeps
A predictive framework is developed to forecast catastrophic landslides by detecting signatures typical of the emergence of dragon‐kings
A phase diagram characterizes the detectability of dragon‐kings against black‐swans and discriminates catastrophic and slow landslides
Abstract
Past exploration missions have revealed that the lunar topography is eroded through mass wasting processes such as rockfalls and other types of landslides, similar to Earth. We have analyzed ...an archive of more than 2 million high-resolution images using an AI and big data-driven approach and created the first global map of 136.610 lunar rockfall events. Using this map, we show that mass wasting is primarily driven by impacts and impact-induced fracture networks. We further identify a large number of currently unknown rockfall clusters, potentially revealing regions of recent seismic activity. Our observations show that the oldest, pre-Nectarian topography still hosts rockfalls, indicating that its erosion has been active throughout the late Copernican age and likely continues today. Our findings have important implications for the estimation of the Moon’s erosional state and other airless bodies as well as for the understanding of the topographic evolution of planetary surfaces in general.
In this paper, we describe the investigations and actions taken to reduce risk and prevent casualties from a catastrophic 210,000 m
3
rockslope failure, which occurred near the village of Preonzo in ...the Swiss Alps on May 15, 2012. We describe the geological predisposition and displacement history before and during the accelerated creep stage as well as the development and operation of an efficient early warning system. The failure of May 15, 2012, occurred from a large and retrogressive instability in gneisses and amphibolites with a total volume of about 350,000 m
3
, which formed an alpine meadow 1250 m above the valley floor. About 140,000 m
3
of unstable rock mass remained in place and might collapse partially or completely in the future. The instability showed clearly visible signs of movements along a tension crack since 1989 and accelerated creep with significant hydromechanical forcing since about 2006. Because the active rockslide at Preonzo threatened a large industrial facility and important transport routes located directly at the toe of the slope, an early warning system was installed in 2010. The thresholds for prealarm, general public alarm, and evacuation were derived from crack meter and total station monitoring data covering a period of about 10 years, supplemented with information from past failure events with similar predisposition. These thresholds were successfully applied to evacuate the industrial facility and to close important roads a few days before the catastrophic slope failure of May 15, 2012. The rock slope failure occurred in two events, exposing a compound rupture plane dipping 42° and generating deposits in the midslope portion with a travel angle of 39°. Three hours after the second rockslide, the fresh deposits became reactivated in a devastating debris avalanche that reached the foot of the slope but did not destroy any infrastructure. The final run-out distance of this combined rock collapse–debris avalanche corresponded to the predictions made in the year 2004.
We use multitemporal analyses based on Synthetic Aperture Radar differential interferometry (DInSAR) to study the slope adjacent to the large Punatsangchhu-I hydropower plant, a concrete gravity dam ...under construction in Bhutan since 2009. Several slope failures affected the site since 2013, probably as a consequence of toe undercutting of a previously unrecognised active landslide. Our results indicate that downslope displacement, likely related to the natural instability, was already visible in 2007 on various sectors of the entire valley flank. Moreover, the area with active displacements impinging on the dam site has continuously increased in size since 2007 and into 2018, even though stabilization measures have been implemented since 2013. Stabilisation measures currently only focus on a small portion of the slope, however, the unstable area is larger than previously evaluated. Highly damaged rock is present across many areas of the entire valley flank, indicating that the volumes involved may be orders of magnitude higher than the area on which stabilisation efforts have been concentrated after the 2013 failure. The results highlight that satellite-based DInSAR could be systematically used to support decision making processes in the different phases of a complex hydropower project, from the feasibility study, to the dam site selection and construction phase.
Seismic data analysis is a powerful tool for remote characterization of rock slope failures. Here we develop quantitative estimates of fundamental rockslide properties (e.g., volume) based solely on ...data from an existing regional seismic network. We assembled a data set of twenty known rockslides in the central Alps (with volumes between 1,000 and 2,000,000 m3) and analyzed their corresponding seismograms. Common signal characteristics include emergent onsets, slowly decaying tails, and a triangular spectrogram shape. The main component of seismic energy is contained in frequencies below ∼3–4 Hz, while higher‐frequency signals may be caused by block impacts. Location estimates were generated using automatic arrival time picks and resulted in a mean location error of 10.9 km. A linear relationship for the detection limit of a rockslide as a function of volume was identified for our seismic station network. To estimate rockslide volume, runout distance, drop height, potential energy, and Fahrböschung (angle of reach), we extracted five simple metrics from each seismogram: signal duration, peak value of the ground velocity envelope, velocity envelope area, risetime, and average ground velocity. Using multivariate linear regression, the combination of duration, peak envelope velocity, and envelope area best estimated event parameters, with r2 values ranging between 0.8 and 0.88. Three new rockslides were then used to validate our method, and volume, runout, drop height, and potential energy were estimated within the correct order of magnitude. When provided with a suitable data set of rockslide events, our method can be easily adapted to other regions and seismic networks.
Key Points
Rockslide seismic signals have a common characteristic signature
Seismic metrics can be used to estimate rockslide parameters
Parameter estimation works for unknown rockslide events in the same region
We leverage on optical and radar remote sensing data acquired from the European Space Agency (ESA) Sentinels to monitor the surface deformation evolution on a large and very active instability ...located in the Swiss Alps, i.e., the Moosfluh rock slope. In the late summer 2016, a sudden acceleration was reported at this location, with surface velocity rates passing from maximum values of 0.2 cm/day to 80 cm/day. A dense pattern of uphill-facing scarps and tension cracks formed within the instability and rock fall activity started to become very pronounced. This evolution of the rock mass may suggest that the most active portion of the slope could fail catastrophically. Here we discuss advantages and limitations of the use of spaceborne methods for hazard analyses and early warning by using the ESA Sentinels, and show that in critical scenarios they are often not sufficient to reliably interpret the evolution of surface deformation. The insights obtained from this case study are relevant for similar scenarios in the Alps and elsewhere.
Water infiltration into fractures is ubiquitous in crustal rocks. However, little is known about how such a progressive wetting process affects fracture stiffness and seismic wave propagation, which ...are highly relevant for characterizing fracture systems in situ. We study the acousto‐mechanical behavior of a free‐standing fractured granite subjected to gradual water infiltration with a downward‐moving wetting front over 12 days. We observe significant differences (i.e., by an order of magnitude) in wave amplitudes across the fractured granite compared to an intact granite, with both cases showing a strong correlation between wave amplitudes and wetting front movement. Effects of water infiltration into the fracture and surrounding matrix on seismic attenuation are captured by a numerical model with parameters constrained by experimental data. Back‐calculated fracture stiffness decreases exponentially with the wetting front migration along the fracture. We propose that moisture‐induced matrix expansion around the fracture increases asperity mismatch, leading to reduced fracture stiffness.
Plain Language Summary
In the shallow layers of the Earth, hydrological cycles such as snowmelt, fog, dew, and rain have been shown to change the moisture content of crustal rocks, which can alter the elastic properties of natural fractures and affect the propagation of seismic waves. Understanding how seismic waves propagate in the near‐surface environment is crucial for the assessment of earthquake hazards and the characterization of geologic heterogeneities. In this work, we perform well‐controlled laboratory experiments to study the acousto‐mechanical behavior of a single fracture in granitic rock subjected to progressive wetting over 12 days. We report that the fracture stiffness decreases exponentially as the wetting front advances along the fracture. Our research sheds light on an important question in fracture characterization: how elastic waves propagate across a fracture undergoing moisture‐induced expansion.
Key Points
A laboratory study establishes a relationship among water imbibition, seismic attenuation, and stiffness evolution in a wetted fracture
Wave amplitudes across a fracture correlate strongly with the wetting front movement of infiltrated water within the fracture
Fracture stiffness exponentially decreases with the advance of the wetting front along the fracture