Abstract The exhumation of Alpine Corsica is the result of Oligocene‐Miocene lithospheric back‐arc extension in the wake of the eastward migration of Apenninic subduction. Although the geometry, kinematics and timing of ductile detachments are well constrained, the brittle extension leading to final exhumation of Alpine Corsica is less documented. We integrate structural observations with stable isotope (δ 18 O–δ 13 C) geochemistry and U‐Pb geochronology of carbonates in fault rocks to constrain the spatial and temporal evolution of the W‐dipping Patrimonio Fault System (PFS). The PFS juxtaposes the Low Pressure/Low Temperature Nebbio units and the syn‐extensional Miocene sedimentary sequences onto the High Pressure/Low Temperature Schistes Lustrés. The PFS initiated in early Miocene (≥20 Ma), as km‐thick top‐to‐W/WNW brittle–ductile zone with pervasive SCC′ fabric and pseudotachylytes, which exploited inherited contractional‐related fabrics on the western limb of the Castagniccia‐Cap Corse antiform. Early distributed extension evolved into more localized sinistral transtension, as recorded by lineated slickensides defining the Patrimonio Fault and by cockade‐bearing splay faults in the footwall. This stage of sinistral transtension, dated between 20 and 14 Ma, was linked to the anticlockwise rotation of the Sardinia‐Corsica block during the opening of the Ligurian‐Provençal basin. A later (≤14 Ma) mid‐ to late‐Miocene extensional faulting, driven by Tyrrhenian‐related tectonics, produced the mature Patrimonio Fault, characterized by multiple anastomosing strands of foliated gouges and phyllonites accommodating ∼2–3 km of displacement. We show that lithospheric back‐arc extension in Alpine Corsica was accommodated by major conjugate E‐ and W‐dipping extensional to transtensional semi‐brittle and brittle fault systems during the Miocene.

Active slip partitioning between the subduction megathrust and the upper plate is investigated in the oblique-convergence setting of the Nazca-South American plate boundary between 33° and 47° S. This segment has two major along-strike bends: the Maipo Orocline (~34° S) and the Arauco Peninsula (~38° S), whereas south of 38° S, lies the intra-arc Liquiñe-Ofqui Fault System (LOFS). Here we examine long- and short-term upper-plate deformation by combining a harmonized catalog of about 2,300 fault-slip measurements in the forearc and arc regions, from which we derive P-T axes using kinematic inversions, along with an integrated seismological database for upper-plate events (1976-2025), including global and local networks. These data are categorized by forearc, arc, and back-arc regions. We identify four distinct tectonic segments based on the spatial distribution of P and T axes in the long-and short-term: (1) 33°-34° S, showing both ~E-W and ~N-S subhorizontal shortening in the forearc and arc areas, suggesting active radial shortening; (2) 34°-37° S, dominated by mostly blind, seismogenic, margin-parallel dextral faults along with NW- and NE-trending structures running at a high angle with respect to the plate margin (called transverse faults here); (3) 37°- 41° S, where margin-orthogonal subhorizontal shortening in the submerged forearc coexists with nearly margin-parallel shortening in the emerged forearc, with contemporaneous dextral slip along the LOFS and ~E-W shortening accommodated by transverse NE-trending dextral and NW-trending sinistral seismically active faults; and (4) 41°-47° S, a region governed mainly by the geometry and kinematics of the LOFS strike-slip duplex. Kinematic indicators on mesoscopic faults (mostly slickenfibers) and a lack of pseudotachylytes suggest a considerable aseismic component to upper-plate fault slip, implying that morphotectonic slip rates may overestimate seismic hazard. Future detailed geodetic data may help better constrain the relative contributions of aseismic and seismic slip on the upper-plate faults in this Andean segment.

Earthquake swarms commonly occur in upper-crustal hydrothermal-magmatic systems and activate mesh-like fault-fracture networks at zone of fault complexity. How these networks develop through space and time along seismic faults is poorly constrained in the geological record. Here, we describe a spatially dense array of small-displacement (< 1.5 m) epidote-rich fault-veins within granitoids, occurring at the intersections of subsidiary faults with the exhumed seismogenic Bolfin Fault Zone (Atacama Fault System, Northern Chile). Epidote faulting and veining occurred at 3-7 km depth and 200-300 °C ambient temperature. At distance ≤ 1 cm to fault-veins, the magmatic quartz of the wall-rock shows (i) thin (<10- µm-thick) interlaced deformation lamellae, and (ii) crosscutting quartz-healed veinlets. The epidote-rich fault-veins (i) include clasts of deformed magmatic quartz, with deformation lamellae and quartz-healed veinlets, and (ii) record cyclic events of extensional-to-hybrid veining and either aseismic and seismic shearing. Deformation of the wall-rock quartz is interpreted to record the large stress perturbations associated with the rupture propagation of small earthquakes. Instead, dilation and shearing forming the epidote-rich fault-veins are interpreted to record the later development of a mature and hydraulically-connected fault-fracture system. In this latter stage, the fault-fracture system cyclically ruptured due to fluid pressure fluctuations, possibly correlated with swarm-like earthquake sequences.

Fluid infiltration along seismically-active faults and fluid-rock interaction influence the mechanical behavior of faults. Nevertheless, how fluid infiltration and fluid-rock interactions evolve at seismogenic depths with fault slip accumulation remain poorly constrained in the geological record. We used hydrogen and oxygen isotope geochemistry to determine the origin of hydrous fluids that percolated within the exhumed Bolfin Fault Zone (BFZ)-a segment of the Early Cretaceous intra-arc Atacama Fault System (Northern Chile)-during progressive fault evolution at seismogenic depth. The BFZ consists of D1 pseudotachylyte-bearing cataclastic strands linked by D2 extensional to hybrid extensional-shear, epidote-rich fault-vein systems that formed in a fluid-rich, seismically active environment at 3-7 km depth and 200-300 degrees C. The D1 pseudotachylytes and cataclasites have delta D values similar to, or slightly higher than, those of unaltered hydrogen-bearing magmatic minerals (-78 parts per thousand <= delta D <= -56 parts per thousand). This similarity indicates that seismic faulting occurred in a rock-buffered environment with limited circulation of external fluids at early stages of fault evolution. Conversely, the epidote of the D2 fault-vein systems has much heavier delta D compositions (-47 parts per thousand <= delta D <= -9 parts per thousand) and delta 18O values ranging from 3.77 to 6.71 parts per thousand, suggesting infiltration of shallow fluids, likely sourced from closed, marine-connected basins. Epidote-quartz oxygen isotope thermometry indicates equilibration at 200-220 degrees C for this stage of fluid infiltration. The influx of external, basin-derived fluids within the BFZ is interpreted to indicate the increased hydraulic connectivity during slip accumulation and fault network growth.

Alpine Corsica is an accretionary wedge formed during the Alpine orogenesis and exhumed through Oligocene-Miocene lithospheric extension controlled by the eastward migration of Apenninic subduction. Here we integrate field geological surveys with microstructural and carbonate stable isotope (δ18O–δ13C) analyses of fault zone rocks to constrain the evolution of the W-dipping extensional Patrimonio Fault System (PFS). The PFS consists of multiple gouge-bearing fault core strands and splay faults in the footwall damage zone, and exhumed the Schistes Lustrés (e.g., impure quartzites, marbles, calcschists, serpentines) and slices of Hercynian granitoids in the footwall block, accommodating ~6 km of cumulative displacement. We describe a deformation sequence during exhumation consisting of D1 mylonitic shearing, D2 seismogenic faulting and D3 shallow veining events. D1 mylonitic shearing produced a decameter mylonitic zone forming the roots of PFS, coeval with localized brittle-ductile shear zones and quartz ± chlorite vein arrays observed in the footwall metamorphic units. Ductile shearing was accommodated by low-temperature quartz and calcite crystal-plasticity, and pressure-solution mechanisms at greenschist conditions (i.e., 300-400 °C). D2 seismogenic faulting either overprinted or cut the D1 structures. Ancient seismic faulting is attested by occurrence of (i) altered pseudotachylytes and (ii) cockade-bearing fault-veins injecting into the host-rocks and mutually overprinting dolomite-rich veinlet mesh and mirror-like slip surfaces observed in the footwall splay faults. Seismic faulting is also accommodated by dolomite-quartz(-chalcedony) crack-seal veins, which have isotopic compositions similar to those of the carbonate-rich units of the Schistes Lustrés. These structural and geochemical observations indicate that ancient seismicity was cyclically modulated by overpressured fluids which isotopic composition was buffered by the host-rocks. The later D3 shallow (≤ 1 km depth) veining event consists of calcite-bearing veins and concretions filling open fractures, which have distinct isotopic compositions compared to the Schistes Lustrés units, suggesting percolation of meteoric fluids at depths. Based on these observations, we speculate that the D2 faults may represent a fossil analogue of the extensional faults active in the Apennines where seismicity is driven by CO2-rich deep-sourced fluids.

To understand better the development of deformation in carbonate-hosted normal faults, we compared the structural architecture of the Campo Felice and Monte D'Ocre active faults (Italian central Apennines). The two geometrically linked structures displace the same carbonate sequences, but with different Quaternary slip rates and geological throws. Moreover, several geomorphological features typical of deep-seated landslides were identified across the Mt. D'Ocre range. The Campo Felice fault segment and the Cama fault segment (Monte D'Ocre range) consist of 0.4-15 m thick and almost absent fault cores and of >400 m and <40 m thick damage zones, respectively. The associated slip zones have different fabrics (i.e., cataclasite vs. crush fault breccia for Campo Felice and Cama Fault, respectively). The different fault zone architecture and associated landscapes would suggest different behaviors of the two faults although similar deformation mechanisms (i.e., cataclasis and pressure-solution) are active in both the two scarps. The Mt. D'Ocre faults would not be segments of the Ovindoli-L'Aquila Fault System and currently accommodate the lateral spreading of the Mt. D'Ocre ridge. Therefore, the seismic hazard associated with the fault system might be reduced. This work shows how macro-to micro -structural analyses provide further information to improve the structural characterization of seismogenic sources.

Faults can act as conduits for the migration of hydrothermal fluids in the crust, affecting its mechanical behaviour and possibly leading to earthquake swarm activity. To date, there are still few constraints from the geological record on how fault-vein networks develop through time in high fluid-flux tectonic settings. Here, we describe small displacement (<1.5 m) epidote-rich fault-vein networks cutting granitoids in the exhumed Bolfin Fault Zone (Atacama Fault System, Chile). The epidote-rich sheared veins show lineated slickensides with scattered orientations and occur at the intersections with subsidiary structures in the fault damage zone. FEG-SEM cathodoluminescence (CL) reveals that magmatic quartz close to the sheared epidote-rich veins is affected by (i) thin (< 10 µm) interlaced deformation lamellae and (ii) a network of CL-dark quartz epitaxial veinlets sharply crosscutting the lamellae. EBSD maps of the deformed quartz indicate minor lattice distortion associated with the lamellae and an orientation nearly orthogonal to the c-axis. These deformation features disappear moving away into the host rock. The epidote-rich sheared veins (i) include clasts of magmatic quartz with both the deformation lamellae and the healed veinlets and (ii) show cyclic events of extensional-to-hybrid veining and localized shearing. We propose that the microstructures preserved in the quartz next to the sheared veins (i.e. deformation lamellae and epitaxial veinlets) record the high-strain rate loading associated with dynamic crack propagation and rapid micro-fracture sealing. On the other hand, the cyclic dilation and shearing within the epidote-rich veins is interpreted as the expression of a highly connected fault-vein network dominated by pore pressure oscillations leading to seismic swarm activity.

Fault zone architecture controls, for instance, the nucleation, propagation and arrest of individual seismic ruptures, the moment magnitude of the mainshocks and the evolution in space and time of foreshock and aftershock seismic sequences. Nevertheless, the architecture of crustal-scale seismogenic sources is still poorly known. Here, we examine the architecture of the >40-km-long, Mesozoic seismogenic Bolfin Fault Zone (BFZ) of the Atacama Fault System (Northern Chile). The exceptionally well-exposed BFZ cuts through plutonic rocks of the Coastal Cordillera and was seismically active at 5-7 km depth and ≤ 300 °C in a fluid-rich environment. The BFZ includes multiple fault core strands consisting of chlorite-rich cataclasites-ultracataclasites and pseudotachylytes, surrounded by chlorite-rich protobreccias to protocataclasites over a zone as wide as 75 m. These cataclastic units are associated with a damage zone, up to 150-m-thick, which comprises strongly altered and brecciated rock volumes, and with clusters of epidote-rich fault-vein networks located at the linkage of the BFZ with other faults. The architecture of the BFZ is the result of fault core widening by cyclic co-seismic frictional melting and post-to-inter-seismic fault healing due to hydrothermal (chlorite + epidote ± K-feldspar) mineral precipitation plus pervasive, possibly associated with mainshocks and aftershocks, damaging of the surrounding rocks. Additionally, we interpret the epidote-rich fault-vein networks as an exhumed seismic source of fluid-driven earthquake swarm-type sequences in agreement with seismological observations of presently active magmatic and hydrothermal regions.

Fault zone architecture and its internal structural variability play a pivotal role in earthquake mechanics, by controlling, for instance, the nucleation, propagation and arrest of individual seismic ruptures and the evolution in space and time of foreshock and aftershock seismic sequences. Nevertheless, the along-strike architectural variability of crustal-scale seismogenic sources over regional distances is still poorly investigated. Here, we describe the architectural variability of the >40-km-long exhumed, seismogenic Bolfin Fault Zone (BFZ) of the intra-arc Atacama Fault System (Northern Chile). The BFZ cuts through plutonic rocks of the Mesozoic Coastal Cordillera and was seismically active at 5-7 km depth and <= 300 degrees C in a fluid-rich environment. The BFZ in-cludes multiple altered fault core strands, consisting of chlorite-rich cataclasites-ultracataclasites and pseudo-tachylytes, surrounded by chlorite-rich protobreccias to protocataclasites over a zone up to 60-m-thick. These fault rocks are embedded within a low-strain damage zone, up to 150-m-thick, which includes strongly altered volumes of dilatational hydrothermal breccias and clusters of epidote-rich fault-vein networks at the linkage of the BFZ with subsidiary faults. The strong hydrothermal alteration of rocks along both the fault core and the damage zone attests to an extensive percolation of fluids across all the elements of the structural network during the activity of the entire fault zone. In particular, we interpret the epidote-rich fault-vein networks and associated breccias as an exhumed example of upper-crustal fluid-driven earthquake swarms, similar to the presently active intra-arc Liquin similar to e-Ofqui Fault System (Southern Andean Volcanic Zone, Chile).

The nucleation and evolution of major crustal-scale seismogenic faults in the crystalline basement as well as the process of strain localization represent a long-standing, but poorly understood, issue in structural geology and fault mechanics. Here, we addressed the spatio-temporal evolution of the Bolfin Fault Zone (BFZ), a >40-km-long exhumed seismogenic splay fault of the 1000-km-long strike-slip Atacama Fault System. The BFZ has a sinuous fault trace across the Mesozoic magmatic arc of the Coastal Cordillera (Northern Chile). Seismic faulting occurred at 5-7 km depth and ≤ 270 °C in a fluid-rich environment as recorded by extensive propylitic alteration and epidote-chlorite veining. The ancient (125-118 Ma) seismicity is attested by the widespread occurrence of pseudotachylytes both in the fault core and in the damage zone. Field geological surveys indicate nucleation of the BFZ on precursory geometrical anisotropies represented by magmatic foliation of plutons (northern and central segments) and andesitic dyke swarms (southern segment) within the heterogeneous crystalline basement. Faulting exploited the segments of precursory anisotropies that were favorably oriented with respect to the long-term stress field associated with the oblique ancient subduction. The large-scale sinuous geometry of the BFZ may result from linkage of these anisotropy-pinned segments during fault growth. This evolution may provide a model to explain the complex fault pattern of the crustal-scale Atacama Fault System.

Abstract Tectonic pseudotachylytes are thought to be unique to certain water‐deficient seismogenic environments and their presence is considered to be rare in the geological record. Here, we present field and experimental evidence that frictional melting can occur in hydrothermal fluid‐rich faults hosted in the continental crust. Pseudotachylytes were found in the >40 km‐long Bolfín Fault Zone of the Atacama Fault System, within two ca. 1 m‐thick (ultra)cataclastic strands hosted in a damage‐zone made of chlorite‐epidote‐rich hydrothermally altered tonalite. This alteration state indicates that hydrothermal fluids were active during the fault development. Pseudotachylytes, characterized by presenting amygdales, cut and are cut by chlorite‐, epidote‐ and calcite‐bearing veins. In turn, crosscutting relationship with the hydrothermal veins indicates pseudotachylytes were formed during this period of fluid activity. Rotary shear experiments conducted on bare surfaces of hydrothermally altered rocks at seismic slip velocities (3 m s−1) resulted in the production of vesiculated pseudotachylytes both at dry and water‐pressurized conditions, with melt lubrication as the primary mechanism for fault dynamic weakening. The presented evidence challenges the common hypothesis that pseudotachylytes are limited to fluid‐deficient environments, and gives insights into the ancient seismic activity of the system. Both field observations and experimental evidence, indicate that pseudotachylytes may easily be produced in hydrothermal environments, and could be a common co‐seismic fault product. Consequently, melt lubrication could be considered one of the most efficient seismic dynamic weakening mechanisms in crystalline basement rocks of the continental crust.

Pseudotachylytes (solidified friction melts produced during seismic slip) are considered to be rare in the geological record because they should be typical of particular seismogenic environments characterized by water-deficient cohesive rocks. Here we present field and experimental evidence that frictional melting can occur in “fluid-rich” faults hosted in the continental crust. Pseudotachylytes were found in the >40 km long Bolfin Fault Zone of the Atacama Fault System (Northern Chile). The pseudotachylytes (1) are associated with a ~1 m thick ultracataclastic fault core which accommodated > 5 km of strike-slip displacement at 6-8 km depth and 280-350°C ambient temperature, (2) cut a ca. 50 m thick damage zone made of sub-greenschists facies hydrothermally altered diorites and gabbros, (3) cut and are cut by epidote+chlorite+calcite bearing veins. The microstructure of the pseudotachylytes include (1) tabular microlites of feldspar hosted in a glassy-like matrix and (2) vesicles filled by post-seismic sub-greenschist facies minerals hosted in a strongly altered matrix of albite, chlorite, and epidote crystals. Experiments reproducing seismic slip in the presence of pressurized water and conducted with the rotary shear apparatus SHIVA on experimental faults made of the sub-greenschists (hydrothermally altered) facies damage zone rocks from the Bolfin Fault Zone, resulted in the production of vesiculated pseudotachylytes. In these experiments, fault weakening mainly occurred by melt lubrication rather than by pore fluid thermal pressurization. The identification of pseudotachylytes and its association with intense pre- and post-seismic hydrothermal alteration challenges the common belief that pseudotachylytes are rare. Consistent with the experimental evidence, pseudotachylytes (1) could be a common coseismic fault product in the continental crust, (2) may easily be produced in fluid-rich hydrothermal environments but, (3) are easily lost from the geological record because they are prone to alteration.

Few fault rocks are known to be associated undoubtedly with seismic faulting. Here, we investigated the formation mechanism of cockade breccias found in transtensional faults cutting marbles and quartzites from the Col de Teghime area (Alpine Corsica, France). Field surveys coupled with detailed microanalytical investigations indicated that: (i) the core clasts of the cockades are composed of host rock fragments >310 μm in size that are suspended in the slipping zones and arranged in inverse grading; (ii) the concentric rims of the cockades show a cyclic zoning made of saddle dolomite + Mg-calcite + goethite + anatase; (iii) the cockade-bearing veins are associated with minor fault veins filled with fine fragments (<300 μm in size) cemented by the same minerals of the cockade rims.

It is well-known that fluid migration in the Earth’s upper crust is strongly controlled by the structure of fault zones. Importantly, fluid migration and pore fluid pressure variations in fault networks control the nucleation and evolution of earthquake sequences. Vein filling in fault zones is one of the most impressive geological signatures of the interaction between fluids and fault zone rocks. A relatively common fault vein filling fabric is the spectacular cockade breccia, consisting of fragments of wall- and fault rocks rimmed and sealed by concentric layers of fluid-precipitated minerals. Consequently, the formation of cockade breccia requires rock fragmentation and cementation in the presence of fluids under particular physical and chemical conditions that may occur during different phases of the seismic cycle. This thesis discusses the structure and the mechanism of formation of cockade breccia hosted in the slipping zones of the Miocene in age extensional brittle faults that cut quartzites and impure dolomitic marbles of the Schistes Lustres Complex from Alpine Corsica (France). Original structural geology field surveys and detailed microstructural (optical cathodoluminescence and scanning electron microscopy; micro-tomography; image analysis) and mineralogical/geochemical (micro-Raman spectroscopy, X-ray powder diffraction, Energy-dispersive X-ray spectroscopy) investigations of the fault rocks indicated that: (a) core clasts of the cockades derive from the wall rocks, have rounded shape and are well-sorted with the fraction finer < 310 µm in diameter almost completely absent; (b) the core clasts of the cockades are suspended (i.e., do not touch each other) in the slipping zones; (c) in some slipping zones, the core clasts are arranged in inverse grading; (d) the concentric layers (systematically four in total) rimming the clasts of the cockades have strong mineralogical zoning consisting of alternate/rhythmic precipitation of saddle dolomite, Mg-calcite and Fe- and Ti-oxides/hydroxides; (e) cockade-breccia are cut but also kinematically associated with veins made of ultrafine (size < 200 µm) wall rock clasts cemented by the same mineral assemblage of the rims of the cockades; (f) the cockade breccia are cut by dolomite-bearing veins and partly sealed by late precipitation of calcite. The above findings allowed me to propose the following model for the formation of the cockade-bearing faults. The model links the formation of the cockade microstructures numbered (a) to (f) to different phases of the seismic cycle: (1) co-seismic fragmentation of the wall rocks (a) in presence of CO2- and Fe-rich fluids which promoted also the rounding of the clasts (abrasion and chemical wear); (2) co-seismic fluidization of the rock fragments associated to fluid pulses migrating in the fault zone. Fluidization resulted in elutriation of the fine particles, which were deposited in distal veins (e), and formation of a residual very porous and well-sorted clast assemblage (a) which will make the core of the cockades. Inverse grading (c) and rounded shape (a) of the cores resulted by shaking (Brazil-Nut Effect) and co-seismic shearing of the clasts; (3) post-seismic to interseismic cementation by deposition of concentric carbonate-rich rims (d) around the core clasts of the cockades. Rim deposition was probably due to slow (years to centuries?) mineral pressure growth processes associated to the ingression of fluids with variable composition in the porous clast assemblage. Pressure growth resulted in the progressive lift of the clasts and in their suspension in the cockade assemblage (b). The precipitation of saddle dolomite and the late and partial sealing of the cockade breccia by calcite cement (f) suggest that the cockade breccia formed at shallow depths in the crust (< 2 km). Based on this conceptual model, cockade breccias are particular fault rock assemblages which record the passage of seismic ruptures in the presence of pressurized migrating fluids. Given the scarcity in the current literature of fault rock assemblages possibly associated to seismic faulting, the results of this study may allow us a better comprehension of earthquake-related processes at shallow crustal depths and find application in seismic hazard studies