Search

International audience; The editors of JGR: Solid Earth announce that Plain Language Summaries will be required for all manuscripts.

International audience; Faults form dense, complex multi‐scale networks generally featuring a master fault and myriads of smaller‐scale faults and fractures off its trace, often referred to as damage. Quantification of the architecture of these complex networks is critical to understanding fault and earthquake mechanics. Commonly, faults are mapped manually in the field or from optical images and topographic data through the recognition of the specific curvilinear traces they form at the ground surface. However, manual mapping is time‐consuming, which limits our capacity to produce complete representations and measurements of the fault networks. To overcome this problem, we have adopted a machine learning approach, namely a U‐Net Convolutional Neural Network, to automate the identification and mapping of fractures and faults in optical images and topographic data. Intentionally, we trained the CNN with a moderate amount of manually created fracture and fault maps of low resolution and basic quality, extracted from one type of optical images (standard camera photographs of the ground surface). Based on a number of performance tests, we select the best performing model, MRef, and demonstrate its capacity to predict fractures and faults accurately in image data of various types and resolutions (ground photographs, drone and satellite images and topographic data). MRef exhibits good generalization capacities, making it a viable tool for fast and accurate mapping of fracture and fault networks in image and topographic data. The MRef model can thus be used to analyze fault organization, geometry, and statistics at various scales, key information to understand fault and earthquake mechanics.

International audience

Measuring fault offsets preserved at the ground surface is of primary importance to recover earthquake and long‐term slip distributions and understand fault mechanics. The recent explosion of high‐resolution topographic data, such as Lidar and photogrammetric digital elevation models, offers an unprecedented opportunity to measure dense collections of fault offsets. We have developed a new Matlab code, 3D_Fault_Offsets, to automate these measurements. In topographic data, 3D_Fault_Offsets mathematically identifies and represents nine of the most prominent geometric characteristics of common sublinear markers along faults (especially strike slip) in 3‐D, such as the streambed (minimum elevation), top, free face and base of channel banks or scarps (minimum Laplacian, maximum gradient, and maximum Laplacian), and ridges (maximum elevation). By calculating best fit lines through the nine point clouds on either side of the fault, the code computes the lateral and vertical offsets between the piercing points of these lines onto the fault plane, providing nine lateral and nine vertical offset measures per marker. Through a Monte Carlo approach, the code calculates the total uncertainty on each offset. It then provides tools to statistically analyze the dense collection of measures and to reconstruct the prefaulted marker geometry in the horizontal and vertical planes. We applied 3D_Fault_Offsets to remeasure previously published offsets across 88 markers on the San Andreas, Owens Valley, and Hope faults. We obtained 5,454 lateral and vertical offset measures. These automatic measures compare well to prior ones, field and remote, while their rich record provides new insights on the preservation of fault displacements in the morphology.

Members of the editorial board of JGR-Solid Earth express their appreciation to those who served as peer reviewers for the journal in 2020.

Earthquake slip distributions are asymmetric along strike, but the reasons for the asymmetry are unknown. We address this question by establishing empirical relations between earthquake slip profiles and fault properties. We analyze the slip distributions of 27 large continental earthquakes in the context of available information on their causative faults, in particular on the directions of their long-term lengthening. We find that the largest slips during each earthquake systematically occurred on that half of the ruptured fault sections most distant from the long-term fault propagating tips, i.e., on the most mature half of the broken fault sections. Meanwhile, slip decreased linearly over most of the rupture length in the direction of long-term fault propagation, i.e., of decreasing structural maturity along strike. We suggest that this earthquake slip asymmetry is governed by along-strike changes in fault properties, including fault zone compliance and fault strength, induced by the evolution of off-fault damage, fault segmentation, and fault planarity with increasing structural maturity. We also find higher rupture speeds in more mature rupture sections, consistent with predicted effects of low-velocity damage zones on rupture dynamics. Since the direction(s) of long-term fault propagation can be determined from geological evidence, it might be possible to anticipate in which direction earthquake slip, once nucleated, may increase, accelerate, and possibly lead to a large earthquake. Our results could thus contribute to earthquake hazard assessment and Earthquake Early Warning.

Knowing the slip amplitudes that large earthquakes produced in prehistorical times is one key to anticipate the magnitude of large forthcoming events. It is long known that the morphology is preserving remnants of paleoearthquake slips in the form of fault-offset landforms. However, the measured offsets that can be attributed to the most recent paleoearthquakes are generally few along a fault, so that they rarely allow recovering the slip distributions and largest slips of these earthquakes. We acquired ~1 m resolution airborne lidar data on a 30 km stretch of a fast-slipping strike-slip fault (eastern Hope Fault, New Zealand) located in a region of high alluvial dynamics where landforms are rapidly evolving. Data analysis reveals >200 offset landforms; only 30% allow a good to moderate quality offset measurement. From these good to moderate quality measures, we recover the slip-length distributions and largest slips of the four most recent large paleoearthquakes and find evidence of 4–6 prior events. The record suggests that large earthquake slip recurred in multiples of about 4 m along the 30 km stretch. Although they have larger uncertainties, the more numerous lower-quality offsets that we also measured reveal a similar earthquake slip record. This shows that, although offset landforms are partly degraded in dynamically active landscapes, they store valuable information on paleoearthquake slips. This information might be recovered provided that the morphology is analyzed at high resolution and " continuously " over a significant fault length. Remote lidar data are powerful to perform such analyses.

We analyze new GPS data spanning 11 years at 92 stations in eastern Iran. We use these data to analyze the present-day kinematics and the slip rates on most seismogenic faults in eastern Iran. The east Lut, west Lut, Kuhbanan, Anar, Dehshir, and Doruneh faults are confirmed as the major faults and are found to currently slip laterally at 5.6 ± 0.6, 4.4 ± 0.4, 3.6 ± 1.3, 2.0 ± 0.7, 1.4 ± 0.9, and 1.3 ± 0.8 mm/yr, respectively. Slip is right-lateral on the ~NS striking east Lut, west Lut, Kuhbanan, Anar, and Dehshir faults and left-lateral on the ~EW Doruneh fault. The ~NS faults slice the eastern Iranian crust into five blocks that are moving northward at 6–13 mm/yr with respect to the stable Afghan crust at the eastern edge of the collision zone. The collective behavior of the ~NS faults might thus allow the Arabian promontory to impinge northward into the Eurasian crust. The ~NS faults achieve additional NS shortening by rotating counterclockwise in the horizontal plane, at current rates up to 0.8°/Ma. Modeling the GPS and available geological data with a block rotation model suggests that the rotations have been going on at a similar rate (1 ± 0.4°/Ma) over the last 12 Ma. We identify large strains at the tips of the rotating east Lut, west Lut, and Kuhbanan faults, which we suspect to be responsible for the important historical and instrumental seismicity in those zones.

We recover the Holocene earthquake history of seven seismogenic normal faults in the Fucino system, central Italy. We collected 800 samples from the well-preserved limestone scarps of the faults and modeled their 36Cl concentrations to derive their seismic exhumation history. We found that > 30 large earthquakes broke the faults in synchrony over the last 12 ka. The seven faults released strain at the same periods of time, 12-9 ka, 5-3 ka, and 1.5-1 ka. On all faults, the strain accumulation and release occurred in 3-6 ka supercycles, each included a 3-5 ka phase of slow (≤ 0.5-2 mm/yr) strain accumulation in relative quiescence, followed by a cluster of three to four large earthquakes or earthquake sequences that released most of the strain in

[1] Past earthquake slips on faults are commonly determined by measuring morphological offsets at current ground surface. Because those offsets might not always be well preserved, we examine whether the first 10 m below ground surface contains relevant information to complement them. We focus on the Te Marua site, New Zealand, where 11 alluvial terraces have been dextrally offset by the Wellington fault. We investigated the site using pseudo-3D Ground Penetrating Radar and also produced a high-resolution digital elevation model (DEM) of the zone to constrain the surface slip record. The GPR data reveal additional information: (1) they image the 3D stratigraphic architecture of the seven youngest terraces and show that they are strath terraces carved into graywacke bedrock. Each strath surface is overlain by 3–5 m of horizontally bedded gravel sheets, including two pronounced and traceable reflectors; (2) thanks to the multilayer architecture, terrace risers and channels are imaged at three depths and their lateral offsets can be measured three to four times, constraining respective offsets and their uncertainties more reliably; and (3) the offsets are better preserved in the subsurface than at the ground surface, likely due to subsequent erosion-deposition on the latter. From surface and subsurface data, we infer that Te Marua has recorded six cumulative offsets of 2.9, 7.6, 18, 23.2, 26, and 31 m (± 1–2 m). Large earthquakes on southern Wellington fault might produce 3–5 m of slip, slightly less than previously proposed. Pseudo-3D GPR thus provides a novel paleoseismological tool to complement and refine surface investigations.

We investigate fault growth in Afar, where normal fault systems are known to be currently growing fast and most are propagating to the northwest. Using digital elevation models, we have examined the cumulative slip distribution along 255 faults with lengths ranging from 0.3 to 60 km. Faults exhibiting the elliptical or “bell-shaped” slip profiles predicted by simple linear elastic fracture mechanics or elastic-plastic theories are rare. Most slip profiles are roughly linear for more than half of their length, with overall slopes always

[1] The Guadeloupe islands are cut by normal faults that accommodate oblique convergence between the North American and Caribbean plates. Such faults are responsible for part of the shallow seismicity and have produced M ≥ 5 damaging earthquakes. To better assess the seismic hazard in Guadeloupe, we quantify the slip rate on one of the largest fault (Morne-Piton). This roughly E-W fault crosses the island of Marie-Galante and uplifts a flight of reef terraces. From geomorphic analysis, we mapped three main terraces. New U/Th datings show that they formed during the latest interglacials, ∼120 and ∼240 kyr ago. Correlation with SPECMAP isotopic records implies that the Marie-Galante plateau emerged during the ∼330 ka highstand. Topographic profiles show that the terraces and the plateau are deformed by faulting. Elastic modeling of their shape constrains the geometry of the fault (70–80° dip, 5 km depth) and its slip rate (0.5 ± 0.2 mm/yr). Given its length (50 km), depth, and slip rate, this fault might produce maximum M ∼ 6.5 earthquakes with a recurrence time of 1400 to 3300 years, or more likely smaller events such as the M ∼ 5.5 16 May 1851 and 3 August 1992 shocks that might recur every 400 to 1000 years. We also show that all the islands and terraces are tilted westward perpendicularly to the trench. La Desirade closest to the trench is uplifted by 276 m, whereas subsidence (−70 m) is observed 10 km east of the volcanic arc. This tilt probably resulted from a transient deformation episode at the subduction interface that predated the late Pleistocene.

[1] We present here geomorphic evidence for an emergent active thrust along the northwestern Apennine front about 50 km south of Milano. Fieldwork combined with SPOT image analysis attests to the presence of a ∼35-km-long, en echelon, cumulative fault scarp, that cuts E-W across Quaternary surface deposits from Casteggio to Sarmato. The scarp offsets vertically alluvial fans and terraces emplaced by tributaries of the Po River that flow northward from the Apennines. Incision by such tributaries occurs only south of the scarp. Valleys of smaller streams stand in hanging position. To the northeast, the buried, east dipping Pavia lateral ramp bounds the San Colombano anticline. We interpret the Montebello-Sarmato scarp to result from the emergence of an active thrust fault dipping south, the Stradella thrust, which splays eastward from that ramp. Total station profiles leveled perpendicular to the Stradella thrust show variable cumulative surface throws ranging from 2 to 25 m. Scarp degradation analysis indicates fairly recent offset of the terrace surfaces (10–100 kyr) suggesting a minimum uplift rate of 0.3 mm/yr. The underlying fault might thus be one of the most active thrust ramps along this segment of the Apennines. N-S to N15°W shortening is consistent with both the E-W strike of this thrust ramp and with coeval oblique sinistral motion on the NNE-SSW trending Pavia lateral ramp. The Vogherese-Bobbiese earthquake of 1828 (I ≈ IX) south of Casteggio and Broni may have activated one segment of this lateral ramp rather than the Stradella thrust. Other historical events (I > 7) that caused damage in Pavia, Piacenza, Lodi, and Milano in the last 2000 years (290 A.D., 1249 A.D., 1276 A.D., 1473 A.D.), may have occurred on the Broni-Stradella thrust or on the blind San Colombano and Lodi thrusts farther north. This makes detailed paleoseismological studies and trenching across its trace mandatory.

[1] Subduction of Atlantic seafloor under the Caribbean plate causes shallow earthquakes within the Lesser Antilles volcanic arc. Such earthquakes, above the subduction interface, show strike-slip or normal fault plane solutions, the latter with ∼E-W striking nodal planes. To better assess seismic hazard and the coupling between volcanism and tectonics, we investigated faulting related to overriding-plate deformation in the Guadeloupe archipelago. Using aerial photographs, satellite SPOT images, and topographic maps (1/25000 scale), we mapped active and middle to late Pleistocene fissures and normal fault systems that cut the uplifted coral platforms Grande-Terre and Marie-Galante and the volcanic rocks of Basse-Terre. The available marine geophysical data show that the faults extend offshore to bound submarine rifts. The E-W striking, 1500 m deep, V-shaped Marie-Galante rift separates the two islands of Marie-Galante and Grande-Terre. Normal faults in the north of Grande-Terre appear to mark the similarly V-shaped, western termination of the 5000 m deep, N°50E to N130°E striking Desirade graben. Three shallow, M ∼ 5.5 earthquakes (6 May 1851, 29 April 1897, 3 August 1992) appear to have ruptured segments of the Marie-Galante rift boundary faults. The young “La Grande Decouverte” volcanic complex of Basse-Terre, including the 1440 A.D. Soufriere dome, lies within the western termination of the Marie-Galante rift. The ancient volcanic shoulders of the rift buttress the active dome to the north and south, which may explain why major prehistoric sector collapses and pyroclastic avalanches have been directed southwestward into the Caribbean Sea, or southeastward into the Atlantic Ocean. The Marie-Galante rift is typical of other troughs transverse to the northeastern edge of the Caribbean plate. We interpret such troughs, which are roughly orthogonal to the arc, to result from slip-partitioning and extension perpendicular to plate convergence. That they disappear southward implies that they result from interaction between the Caribbean and North American plates.