Focused Fluid Flow Research Articles

Hydrothermal alteration, fluid flow and volume change in shear zones: the layered mafic–ultramafic Kettara intrusion (Jebilet Massif, Variscan belt, Morocco)

AbstractDuring emplacement and cooling, the layered mafic–ultramafic Kettara intrusion (Jebilet, Morocco) underwent coeval effects of deformation and pervasive fluid infiltration at the scale of the intrusion. In the zones not affected by deformation, primary minerals (olivine, plagioclase, clinopyroxene) were partially or totally altered into Ca‐amphibole, Mg‐chlorite and CaAl‐silicates. In the zones of active deformation (centimetre‐scale shear zones), focused fluid flow transformed the metacumulates (peridotites and leucogabbros) into ultramylonites where insoluble primary minerals (ilmenite, spinel and apatite) persist in a Ca‐amphibole‐rich matrix. Mass‐balance calculations indicate that shearing was accompanied by up to 200% volume gain; the ultramylonites being enriched in Si, Ca, Mg, and Fe, and depleted in Na and K. The gains in Ca and Mg and losses in Na and K are consistent with fluid flow in the direction of increasing temperature.When the intrusion had cooled to temperatures prevailing in the country rock (lower greenschist facies), deformation was still active along the shear zones. Intense intragranular fracturing in the shear zone walls and subsequent fluid infiltration allowed shear zones to thicken to metre‐scale shear zones with time. The inner parts of the shear zones were transformed into chlorite‐rich ultramylonites. In the shear zone walls, muscovite crystallized at the expense of Ca–Al silicates, while calcite and quartz were deposited in ‘en echelon’ veins. Mass‐balance calculations indicate that formation of the chlorite‐rich shear zones was accompanied by up to 60% volume loss near the centre of the shear zones; the ultramylonites being enriched in Fe and depleted in Si, Ca, Mg, Na and K while the shear zones walls are enriched in K and depleted in Ca and Si. The alteration observed in, and adjacent to the chlorite shear zones is consistent with an upward migrating regional fluid which flows laterally into the shear zone walls. Isotopic (Sr, O) signatures inferred for the fluid indicate it was deeply equilibrated with host lithologies.

Epigenetic Nature of the BIF-hosted Gold Mineralization at Ajjanahalli, Southern India: Evidence from Ore Petrography and Fluid Inclusion Studies

Gold mineralization at Ajjanahalli is BIF-hosted and structurally controlled. Ore petrographic studies revealed that gold was introduced as a consequence of focussed fluid flow and sulfidation of an oxide facies BIF horizon. Ubiquitous syn-kinematic quartz veins within the BIF as well as in the associated metasedimentary units lay testimony to the high fluid flux during deformation. Fluid inclusion studies involving microthermometry and laser Raman microspectroscopy on inclusions from quartz veins from different litho-units were carried out. The primary fluid preserved as aqueous-carbonic inclusions within quartz veins in carbonaceous phyllites is found to be H 2O-CO 2-CH 4-NaCl type. Near-isothermal decompression led to splitting of this fluid at greenschist facies conditions (0.8 kb/263°C to 1.15 kb/309°C) and resulted in a low saline aqueous and a carbonic (+CH 4) fluid. The auriferous aqueous part reacted with the oxidic BIF, leading to epigenetic replacement of oxides to pyrite and concomitant precipitation of gold.

Geomorphological, geophysical, and geochemical evidence of fluid flow through the seabed

Historically, seabed fluid flow (especially groundwater expulsion) has been known to man for centuries. However, modern survey systems such as the side-scan sonar, sub-bottom profiler, and, not the least, the multi-beam echo sounder have facilitated mapping of seabed details at all water depths. Focused fluid flow locations on the seabed are characterised by geomorphological expressions such as pockmarks, carbonate mounds, and mud volcanoes. When these topographical expressions are combined with other geophysical data, such as acoustic reflection characteristics, side-scan sonar data, and shallow seismic data, fluid flow locations can normally be pin-pointed. Ground-truthing of focused fluid flow through the seafloor is performed either visually or by geochemical sampling and analyses.

BASALTIC BRECCIAS IN THE UPPER OCEANIC CRUST, HOLE 504B (COSTA RICA RIFT, PACIFIC OCEAN)

DSDP/ODP Hole 504B is located in the Panama Basin, eastern Pacific ocean, 200 km south of the Costa Rica Rift, which is a 171-km-long east-west trending segment of the Cocos-Nazca Spreading Center. Hole 504B penetrates 2.1 km into 5.9 m.y. old oceanic crust consisting, from the top to the bottom, of pillow lavas (571.5 m) covered by 274.5 m of sediments; a transition zone (209 m), and a lower zone composed of diabasic dikes (1056 m). Basaltic breccia is a significant component of the volcanic section, down to the upper part of the sheeted dike complex. Six main types of breccias were recognized through core and petrographic observations. Only jigsaw-puzzle breccias can be interpreted as tectonic. Although overall pore-fluid pressures slightly lower than hydrostatic were measured at Hole 504B, jigsaw-puzzle breccias may have formed under local conditions of suprahydrostatic fluid pressure, mostly beneath impermeable barriers such as massive basaltic flows. High fluid pressure conditions were favoured by a structural setting allowing the action of a fault-valve. Highly fractured zones identified by core observations and geophysical logs at depths of about 400-550 mbsf and 800-1100 mbsf, respectively, likely correspond to faults, creating the conditions for a fault-valve mechanism. Breccias of the stockwork zone were produced by focused fluid flow associated with a discharge hydrothermal cycle. All the other breccias recognized in the studied section are interpreted as deriving from either in situ fragmentation of basalts or sedimentation of basaltic debris.

Relations between progressive deformation and fluid-rock interaction during shear-zone growth in a basement-cored thrust sheet, Sevier orogenic belt, Utah

Variations in microtextures, strain, whole-rock chemistry, mineralogy, fluid inclusion characteristics, and fracture network properties record complex interactions between deformation and fluid-rock processes during progressive growth of a shear zone within crystalline basement rocks. This shear zone formed at a depth of about 15 kilometers, T ≈ 350°C, and at elevated fluid pressures. The shear zone (SZ) has a 10 meter thick core of highly deformed phyllonite, and is bounded by a 3 to 8 meter thick transition zone (TZ) of variably fractured chloritic gneiss, which grades outward into relatively undeformed granitic gneiss. Granitic gneiss consists mostly of coarser-grained feldspar and quartz; chloritic gneiss consists of varying amounts of quartz, feldspar, and fine-grained micaceous matrix produced by a mixture of cataclastic, plastic, and alteration processes; and phyllonite consists mostly of very fine-grained quartz-mica-rich matrix produced by pervasive plastic deformation and alteration processes. Estimated strain ratios are 10:1 within the SZ. Whole-rock chemical compositions record significant depletion of Ca and Na, and enrichment of Mg and H2O during progressive alteration of granitic gneiss into phyllonite. Similar average contents of Al, Ti, and Fe, as well as Si, indicate that alteration was about isovolumetric at outcrop scale, although variations between individual samples record up to ±20 percent local volume change. Alteration produced significant changes in mineral abundances and compositions, with conversion of feldspar to muscovite, mafic minerals to Mg-rich chlorite, and dissolution/precipitation of quartz. Fluid inclusion characteristics and mineral compositions indicate that SZ fluids were moderately saline, and became depleted in Mg during alteration. Changes in mineralogy and fluid composition record large influxes of reactive fluids, with geochemical fluid-rock ratios on the order of 102 to 103. Variably deformed, cross-cutting veins, fractures, and microcracks record repeated episodes of cataclasis, fluid influx, and sealing along complex networks. Fluid flow appears to have been concentrated within the SZ along grain-scale and vein networks, with a component of outward flow into the TZ along fracture and microcrack networks. Estimated average fluid fluxes are on the order of 10−7 to 10−10 m/s in the SZ, and 10−9 to 10−12 m/s in the TZ, consistent with average permeabilities of 10−14 to 10−17 m2 in the SZ and 10−16 to 10−19 m2 in the TZ for moderate fluid pressure gradients. These average permeabilities are consistent with observed grain-scale and fracture network properties for a range of sealing and fluid pressure histories. Permeabilities may have been transiently greater during episodes of very high fluid pressure, but decreased as fluid pressure gradients equilibrated and fractures sealed. Strain softening in the SZ was likely produced by reaction softening, grain size reduction, and hydrolytic weakening of quartz. A model of shear zone growth involves: (1) initial fracturing of relatively strong, coarser-grained, quartz-feldspar-rich rock; (2) episodic influx of reactive fluids during periods of fracturing and high fluid pressure, with intervening periods of sealing and reduced fluid pressure; (3) progressive alteration of feldspar and mafic minerals to micas and recrystallization of quartz to form relatively weak, fine-grained matrix; and (4) concentrated deformation and focused fluid flow in a growing SZ core.

Post-deformation, sulphide-quartz vein hosted gold ore in the footwall alteration zone of the Palaeoproterozoic Långdal VHMS deposit, Skellefte District, northern Sweden

The Palaeoproterozoic, c. 1.88 Ga old Långdal VHMS deposit is situated in the eastern part of the Skellefte District, northern Sweden. In the stratigraphic footwall to the VHMS ore a sulphide-quartz vein system with high gold grades was mined in the second half of the 1990′s. The Långdal VHMS ore is hosted by the uppermost part of the Skellefte Group volcanic rocks, close to the contact with an overlying fine-grained sedimentary unit. Regional structural studies indicate that bedding surfaces in volcanic rocks are parallel to the contact between the volcanic and the sedimentary rocks. Based on the differences in structural style on each side, the contact is interpreted as a major structural break. The Långdal ore is situated close to this break that may have focussed fluid flow during metamorphism and deformation. The orientation of the contact indicates that it either is a D2 structure or that it was at least active during D2. The structural development in the altered footwall rocks to the Långdal VHMS ore indicates that gold-bearing sulphide and sulphide-quartz veins both pre- and post-date the first deformation. Gold associated with the vein system can thus not only be attributed to syngenetic exhalative or replacement processes. The close spatial relationship with the massive sulphide deposits suggests, however, that the gold was remobilized from these syngenetic systems. It is concluded that sulphides were introduced at several stages during the geological evolution of the area as: a) syngenetic disseminations of sulphide and folded, pre-S1 stringer sulphide±quartz veins in the footwall related to the syngenetic VHMS deposit, b) syn-S1 sulphide veins in the footwall gold ore, c) main, post-S1, sulphide-quartz veins associated with the gold ore in the footwall rocks to the Långdal VHMS deposit, and d) post-S1 to pre-S2 galena and sphalerite rich veins post-dating the main, post-S1, sulphide-quartz vein system in the footwall to the Långdal ore. From these relationships it is suggested that gold was re-mobilized from the sulphide rich parts of the VHMS system into post-D1 structures during or slightly after the peak metamorphism. The timing of this event is poorly constrained to post-date the syngenetic ore emplacement by 20–80 m.y.

Genesis of mud volcanoes in sedimentary basins: A solitary wave‐based mechanism

Mud volcanoes represent the ground surface morphological expression of highly focused fluid flow inside sedimentary basins. Sedimentary basins show usually the presence of overpressured compartments encapsulated by low‐permeability bounds (seals), which can have different origins (capillary barriers with high entry capillary pressure, diagenetically sealed formations..). A mud volcano can be formed in response to the hydrofracturation of such a permeability barrier. When introducing the pressure dependence of permeability in poroelasticity, the theory predicts the upsurge of fluid pressure fronts after the hydrofracturation of the seal (initial shock). These fronts take the form of hydro‐mechanical non‐linear shock waves, which are solutions of a Bürgers‐type equation. These (solitary) waves are associated with an upsurge of pore water, which depends on the pore fluid pressure gradient carried up with the waves. The arrival of these waves at the ground surface can generate mud volcanoes‐type features with spots of highly focused flow.

Multiple fluid flow events and the formation of saddle dolomite: case studies from the Middle Devonian of the Western Canada Sedimentary Basin

In Western Canada Sedimentary Basin (WCSB), large scale and focused fluid flow that caused hydrothermal dolomitization have been suggested with different timing and intensity. In this study, we conducted a petrographic and geochemical comparison between the Middle Devonian Sulphur Point and Slave Point carbonates from northwestern Alberta. The results demonstrate the presence of both an early fluid flow event associated with the Late Devonian to Mississippian Antler Orogeny as well as a later event coincident with the Late Cretaceous to Early Tertiary Laramide Orogeny. Early fluid flow event is characterized by high salinity fluids, and high temperatures and oxygen isotopic values of marine or slightly enriched values, as demonstrated in saddle dolomite from the Slave Point Formation. In contrast, later fluids that caused the precipitation of saddle dolomite in the Sulphur Point formation are characterized by having slightly saline values, comparable homogenization temperatures but more enriched δ18O values and slightly depleted δ13C signatures. Geochemical data indicate that this later hydrothermal fluid was mixture of Middle Devonian brines and radiogenic basement fluids.

Implications of permeability and its anisotropy in a mica gouge for pore pressures in fault zones

Large permeability anisotropy due to anisotropic alignment of clay shape fabric may play an important role in focusing fluid flow along fault zones in clay rich rocks. The degree of permeability anisotropy depends on the differences in tortuosity of the fluid flow path and mean pore aperture, which are of a function of effective stress. We have conducted hydrostatic compaction and frictional sliding experiments on 1 mm thick artificial muscovite gouge layers with grain size <90 μm, at normal stresses on the layer ranging from 25 to 125 MPa. Permeability was measured both perpendicular and parallel to the gouge layer using water as pore fluid. Shear displacement up to 200 mm was achieved. During initial packing and subsequent hydrostatic compaction, muscovite basal planes were aligned parallel to the fault plane, leading to the largest possible contrast in tortuosity between flow path perpendicular and parallel to the fault zone. We found that fluid permeability parallel to the fault zone is about one order of magnitude higher than perpendicular to it and that the magnitude of permeability anisotropy does not change with increasing normal stress. During shear deformation, the permeability anisotropy decreased with increasing shear displacement, an effect interpreted as being due to progressive rotation of muscovite basal planes away from the fault plane. Previous hydrogeological modelling has shown that permeability anisotropy on the order of 3–5 orders of magnitude is needed to effectively focus fluid flow along fault zones. Thus, our results seem to indicate that fabric anisotropy alone is not enough to cause focused fluid flow along natural fault zones. During hydrostatic compaction of muscovite gouge, a linear relationship holds between ln permeability and normal stress and the slope is about 0.03 MPa −1, which is higher than for most porous sandstones. The value is reduced to about 0.006 MPa −1 after shear deformation. The values are lower by 1–2 orders of magnitude than the value used in Rice's fault model, suggesting that maintaining high pore pressures would be harder than he assumed.

A numerical model of deformation and fluid-flow in an evolving thrust wedge

To investigate deformation and fluid-flow in an actively deforming tectonic wedge, we model the evolution of a large, two-dimensional (100 km long, 5 km thick), mechanically and hydrologically homogeneous and isotropic pile of sedimentary strata that is deformed to become a thrust wedge. We compare both ‘dry’ and ‘wet’ cases, in order to illustrate: (1) the relative importance of fluids on wedge evolution, and (2) the effect of brittle deformation on fluid-flow. We use an elastic–plastic constitutive relation, including a Mohr–Coulomb failure criterion and a non-associated flow rule, and coupled fluid flow, with bulk rock properties that approximate typical foreland sedimentary strata. Simulations are made both with and without dilation. The model is fully dynamic, but inertial forces remain small. Results show that deformation within the wedge is accommodated by reverse-slip movement on shear bands, which migrate in both directions through the wedge as both fore- and back-thrusts. The model has features predicted by the critical-taper theory: (1) overall wedge geometry; (2) crudely self-similar growth during evolution; (3) more intense deformation toward the rear of the wedge. The models show strong overall in-sequence faulting behavior with major thrusts isolating relatively undeformed packages, which are moved in a piggyback manner upon the active thrusts. Intermittent out-of-sequence faulting does however occur, in order to maintain the wedge taper. Fluid-flow in the deforming wedge is dominated by topography, but is also strongly affected by dilational plastic deformation. In all simulations, there is focused fluid flow within fault zones. When mechanical time-stepping is shut off (uncoupled), flow systems evolve to steady-state where inflow equals outflow. By subtracting the two ‘states’ we isolate the mechanical fluid response from the total coupled system response. The mechanical fluid response is manifest as contours of head and pressure difference and focusing of fluid-flow in areas of active deformation. In non-dilational simulations, these zones are quite localized and are a result of the contrasting stresses in and adjacent to a brittle shear zone. In simulations with dilation, the zones of head and pressure difference are broad and regional in extent. They are ten times greater in magnitude than in non-dilation simulations, and they track a rock-damage front that moves through the wedge.

Linking of regional and local hydrothermal systems in the mid-crust by shearing and faulting

The concentration of mass during regional metamorphism and metallogenesis requires interactions between regional and local hydrothermal systems that are scale- and time-dependent. During regional low- to medium-temperature metamorphism in which fluid flow is driven by broad thermal or mechanical gradients, fluid flow rates are (on average) <10 −8 m/s, and are limited by restricted permeabilities and gentle gradients in hydraulic head. In contrast, coseismic fluid flow rates in major fault zones may reach ≥10 −2 m/s. As a consequence, there may be a limited capacity for individual faults to substantially control fluid flow in the region because of the large contrast in flow rates and permeabilities between faults and low permeability rocks. To produce a major epigenetic ore deposit in overpressured crust, fluid flow systems with such diverse flow rates must interact in such a way that metal dissolution and transport from regional fluid flow systems culminate in focused fluid flow with metal deposition. This is best achieved through two steps; one in which km-scale hydrothermal systems channel into 100 m-scale ductile shear zones largely by reaction- and deformation-enhanced permeability in the shear zone; and a second one in which rapid fluid pressure changes accompanying brittle faulting (or intense shearing) within the broader shear zones causes metal deposition and ore genesis.

Focused fluid flow along faults in the Monterey Formation, coastal California

Fluid flow in fractured siliceous mudstone of the Miocene Monterey Formation of California is inferred to be highly focused toward map-scale faults that locally contain extensive amounts of carbonate and minor silica cement. The distance of cross-stratigraphic flow, as inferred based on the strontium isotopic composition of carbonate fault cement, is close to the thickness of the Monterey Formation of 700 m in one of two study locations, Jalama Beach, and less than the formation thickness at another location, Arroyo Burro Beach. Fluid is thus derived from within the Monterey Formation rather than from underlying older units. Based on mass-balance estimates of the fluid volume required for fault cementation at Jalama Beach, the minimum distance of formation-parallel flow into the fault zone is 4 km and possibly >12 km. The inferred distance of flow parallel to the formation into this fault thus exceeds the distance of cross-formational upward flow along the fault by at least a factor of six. The mass-balance estimate requires that fluid flow along the fault is channeled into a pipe-shaped conduit rather than distributed along fault strike. Fluid flow from the surrounding formation into fault pipes is inferred to follow a radial rather than uni- or bilateral flow symmetry, using bedding-confined sets of extension fractures and stratabound breccia bodies. Radial fluid flow toward fault pipes requires fairly isotropic fracture permeability for flow along bedding and a low permeability across bedding. The inferred flow geometry illustrates the combined effect of fault permeability structure, permeability anisotropy of the surrounding formation, and hydraulic head distribution in controlling basinal fluid flow in faulted sequences.

Permeability of a decollement zone: Results from a two‐well experiment in the Barbados accretionary complex

Fault zones are hypothesized to play a major role in focusing fluid expulsion from accretionary complexes. A small number of previous investigations have examined the hydrogeologic properties of these fault zones using borehole packer or submersible‐based slug and constant rate flow tests. While these single‐well experiments yielded the first data on the permeability of active fault zones, they could only investigate a small area around the borehole and could not ensure reliable determination of formation storage properties. Recent data from an unplanned experiment in the décollement zone of the Barbados accretionary prism demonstrate the potential for multiple‐well hydrogeologic tests. Pressure data from a sealed borehole show a clear signal of drilling activity ∼45 m away. Analysis of the pressure response suggests that decollement zone permeability lies between 1.0 and 1.2×10−14m2. This permeability is ∼2–4 orders of magnitude greater than overlying and underlying sediment and thus appears sufficient to focus fluid flow along the decollement zone. This inadvertent test evaluated fault zone hydrogeologic parameters over a larger radius and longer timescale and at pressures closer to in situ than previous single‐well tests.

Calcium mass transport and sandstone diagenesis during compaction-driven flow: Stevens Sandstone, San Joaquin basin, California

Sediment compaction provides a limited source of fluids for diagenesis and drives fluid flow at average rates of only ∼1 mm/yr. Nevertheless, many geochemical and petrographic studies of diagenesis provide evidence of significant mass transport in systems where fluid flow appears to be compaction driven. This apparent discrepancy could arise because diagenetic studies generally are concerned with relatively permeable petroleum reservoirs. Focused fluid flow through permeable zones could increase local flow rates and allow mixing of fluids from different sources. This study uses the Stevens Sandstone of the San Joaquin basin to explore the potential diagenetic effects of fluid focusing during compaction-driven flow over a 5 m.y. period. Reactive-transport simulations incorporate a new kinetic expression for plagioclase dissolution and suggest that rate-limited plagioclase dissolution drove calcium enrichment of pore fluids and caused precipitation of calcite, kaolinite, and albite. In spite of the importance of this rate-limited reaction, simulations show that influx of fluids from compacting shale could limit the distribution of calcite and kaolinite in adjacent sandstones. Although calcium mass transport is predicted on a scale of kilometers, upward flow of calcium-rich fluids from deep beds does not significantly increase calcite volume relative to closed-system predictions. Increased transverse dispersivity increases mixing, which further limits precipitation of calcite and kaolinite. Results are consistent with field observations of fluid chemistry, although simulations account for

Barite–Celestine Geochemistry and Environments of Formation

Minerals in the barite (BaSO4)–celestine (SrSO4) solid solution series, (Ba,Sr)SO4, occur in a remarkably diverse range of sedimentary, metamorphic, and igneous geological environments which span geological time from the Early Archean (~3.5 Ga) to the present. The purpose of this chapter is to review: (1) the controls on the chemical and isotopic composition of barite and celestine and (2) the geological environments in which these minerals form. Some health risks are associated with barite, and these are discussed near the end of this chapter. Although complete solid solution exists between BaSO4 and SrSO4 most representatives of the series are either distinctly Ba-rich or Sr-rich. Hence, it is convenient to use the term barite to refer to not only the stochiometric BaSO4 endmember but also to those (Ba,Sr)SO4 solid solutions dominated by Ba. Similarly, the term celestine will refer here not only to the stoichiometric SrSO4 endmember but to solid solutions dominated by Sr. Such usage is in accord with standard mineral nomenclature. The Committee on Mineral Names and Nomenclature of the International Mineralogical Association recognizes “celestine” as the official name for SrSO4. However, the name “celestite” is still commonly used in the literature. ### Geological significance of barite and celestine Most of the barite which exists in the Earth’s crust has formed through the mixing of fluids, one containing Ba leached from silicate minerals, and the other an oxidized shallow fluid, such as seawater, which contains sulfate. Large deposits of barite represent areas of focused fluid flow and mineral precipitation and thus aid in the reconstruction of the hydrogeologic history of the Earth’s crust. The stability of barite is redox sensitive, and the presence or absence of this mineral helps to constrain interpretation of paleoredox conditions. Pelagic barite, the precipitation of which is biologically mediated, may …

Early establishment of seafloor hydrothermal systems during structural extension: paleomagnetic evidence from the Troodos ophiolite, Cyprus

Paleomagnetic data from the Troodos ophiolite are used to help constrain models for the relationship between extensional normal faulting and hydrothermal alteration related to production of large-tonnage sulfide deposits at oceanic ridges. We have sampled dikes from the Troodos sheeted complex that have been subjected to variable hydrothermal alteration, from greenschist alteration typical of the low water/rock mass ratio interactions outside of hydrothermal upflow zones as well as from severely recrystallized rocks (epidosites) altered within high water/rock mass ratio hydrothermal upflow zones in the root zones beneath large sulfide ore deposits. These dikes are moderately to highly tilted from their initial near-vertical orientations due to rotations in the hangingwalls of approximately dike-parallel, oceanic normal faults. Comparison of characteristic remanence directions from these dikes with the Late Cretaceous Troodos reference direction, therefore, allows a tilt test to determine whether remanent magnetizations were acquired prior to or subsequent to tilting. Remanence directions for both greenschist and epidosite dikes show similar magnitudes of tilting due to rotational normal faulting and restore to the Late Cretaceous Troodos reference direction upon restoration of dikes to near-vertical positions about a NNW-trending, horizontal axis. These data, along with field observations of focused alteration along normal faults, suggest that epidosite alteration occurred during the early stages of extensional tilting and prior to significant rotation. This sequence of events is similar to that observed for creation of large-tonnage sulfide bodies at intermediate to slow spreading centers which form soon after cessation of magmatism and during the early stages of structural extension. We suggest that the dike-parallel normal faults were initiated as extensional fractures during this early stage of crustal extension, thus providing the necessary permeability for focused fluid flow, and that later slip along these structures during rotational-planar normal faulting caused reduction in permeability due to gouge formation.

Numerical modelling of focussed fluid flow in theCascadia accretionary wedge

Based on seismic reflection and borehole data (ODP Leg 146) from the Cascadiaaccretionary wedge off Oregon we use the finite element method to model coupled hydraulic andthermal processes in a fault and its vicinity.In a simple model geometry, different boundaryconditions and a wide range of material properties are tested in several model series. The mostreasonable model parameter set includes permeabilities of 5×10−13 m2for the fracture and 5×10−17 m2 for the rock matrix. Calculated flowvelocities in the fracture zone are then at 1.9×10−7 ms−1, which is aboutthree orders of magnitude higher than diffuse flow in the surroundings.To construct a perhaps more realistic model geometry, a transition zone between thefracture and the surrounding rock matrix with intermediate material properties is inserted tosupport the focussing effect on the fluids. A higher transition zone permeability enhances thiseffect. It also influences the magnitude of fluid velocity within the fault. Depending on thepermeability contrast, flow velocities range from 2.8×10−8 ms−1 to1.1×10−7 ms−1. Due to the drop of pore pressure difference caused bythe transition zone flow, a higher permeability in the transition zone reduces the fluid velocity inthe fracture and could subsequently lead to a closure of the fracture.The complex model seems to be closer to reality than the simple one, also providing anarrow range of 1 to 2 m for the fracture width, which could not be determined exactly bymeasurements. Both models show that fractures provide an effective mechanism of dewatering anaccretionary wedge, with tectonically driven flow velocities in the fracture at least three orders ofmagnitude higher than in the surrounding porous medium. Assuming, in accordance with Suess et al., 1998, that vent sites cover about one per mile of the seafloor, this indicates thatdiffuse and focussed fluid flow may contribute nearly equally to dewatering of an accretionarywedge. © 1999 Elsevier Science Ltd. All rights reserved.

Experimentally based water budgets for dehydrating slabs and consequences for arc magma generation

Phase diagrams of hydrous mid-ocean ridge (MOR) basalts to 330 km depth and of hydrous peridotites to 250 km depth are compiled for conditions characteristic for subduction zones. A synthesis of our experimentally determined phase relations of chlorite, lawsonite, epidote-zoisite, amphibole, paragonite, chloritoid, talc, and phengite in basalts and of phase relations from the literature of serpentine, talc, chlorite, amphibole, and phase A in ultramafics permits calculation of H 2O contents in hydrous phase assemblages that occur in natural compositions. This yields the information necessary to calculate water budgets for descending slabs. Starting from low-grade blueschist conditions (10–20 km depth) with H 2O contents between 5 and 6 wt% for hydrated oceanic crust, complete dehydration is achieved between 70 and >300 km depth as a function of individual slab geotherms. Hydrous phases which decompose at depth below volcanic arcs are lawsonite, zoisite, chloritoid, and talc (± phengite) in mafic compositions and chlorite and serpentine in peridotite. Approximately 15–35% of the initially subducted H 2O are released below volcanic arcs. The contribution of amphibole dehydration to the water budget is small (5–20%) and occurs at relatively shallow depth (65–90 km). In any predicted thermal structure, dehydration is a combination of a stepwise and a continuous process through many different reactions which occur simultaneously in the different portions of the descending slab. Such a dehydration characteristic is incompatible with `single phase dehydration models' which focus fluid flow through a unique major dehydration event in order to explain volcanic fronts. As a consequence of continuously progressing dehydration, water ascending from the slab will be generally available to depth of ca. 150–200 km. The fluid rising from the subducting lithosphere will cause partial melting in the hot portion of the mantle wedge. We propose that the volcanic front simply forms above the mantle wedge isotherm where the extent of melting is sufficient to allow for the mechanical extraction of parental arc magmas. Thermal models show that such an isotherm (ca. 1300°C) locates below volcanic fronts, slab surface depths below such an isotherm are compatible with the observed depths of the slab surface below volcanic fronts.

Fracture Cementation in the North Sea

Fluid migration and distribution are important aspects of the evolution of sedimentary basins. An essential parameter affecting fluid flow is permeability. Even mm thick fractures have permeabilit ies many orders of magnitude above surrounding sediments, and thus have the potential of focusing fluid flow. It is a major task to establish the importance of these fractures as fluid conduits. One aspect is to determine the mechanisms of mass transport and precipitation of the fracture cements. Basically the question is whether the mineral cementing fluids are locally sourced from the wall rock by diffusion or local fluid advection, or have been externally imported and focused into the fractures, or both. Detailed investigation of vein mineral and wall rock may help in establishing the relative importance of the two major cementing processes. Here, we will focus on the use of petrographical observations. In addition, we use microscale stable isotope analysis of carbonate cemented fractures obtained with the laser ablation method (O and C) in order to establish the direction of material transfer. Fracture cements. Cemented fractures occur in various lithologies of Tertiary to Palaeozoic ages within the North Sea, including sandstones, mudrocks, chalk, limestones and volcanics. The fractures are commonly subvertical and 1-2 mm thick, rarely exceeding 1-2 cm, and they are in many cases reactivated. The fractures are cemented by a variety of minerals. However, the about 150 fracture occurrences recorded in our North Sea fracture database are dominantly cemented by carbonates. Other minerals include kaoline, baryte, pyrite and only occasional quartz. This observation contrasts with the common occurrence of quartz veins in metamorphic rocks. Bitumen is also sometimes seen as a fracture filling material. Cemented fractures may be monomineralic, or they are seen to be filled with more than one mineral type. Carbonate cement. Calcite, dolomite/ankerite and siderite are all observed. The grain size of the fracture carbonates is in most cases notably coarser than the grain size in wall rock cements.The wall rock cement is generally also carbonate, but chemically distinct from the vein mineral (e.g. siderite versus dolomite or calcite). Multiple generations of carbonate cementation in fractures occur as well, and also precipitation by carbonate preceding a quite different phase as kaoline is occasionally observed. Carbonate fractures in sandstones which are not cemented by carbonates are recorded in a few cases. Pyrite cement. Pyrite fracture cements are less common, but are observed both within sandstones and altered volcanics. Some fractures have walls lined by early pyrite, and central parts cemented by carbonate, thus documenting the presence of compositionally different fluids at various stages of fracture cementation. Kaoline cement. Fractures cemented by kaoline occur within sandstones and interlayered igneous rocks.The texture is fine grained and appear similar to pore-filling, blocky kaoline cement. Wall rock cements recorded are carbonates or clays. Baryte cement. Baryte cemented fractures are observed within all lithologies. The grains are commonly coarse compared to wall rock, which is cemented by either clays or carbonates. Both equant and elongated grains occur. Dendritic growth is recorded in one case. Quartz cement. Quartz cemented fractures are rarely observed and pure quartz veins are only found within altered igneous rocks. These are probably due to hydrothermal activities. Chert veins are seen within chalk, and carbonate fractures hosted in sandstones occasionally contain a few grains of coarse quartz, which evidence the presence of silicate fluids. Isotopes. The laser ablation method allows for very small-scale recording of variations in isotope composition (C, O) across a carbonate filled fracture. One example of such a carbon isotope profile is given in Fig. 1. The carbon isotopes demonstrate that the carbon source for fracture mineralisation was different from that of the local wall rock.If the fracture fillings has been formed by wall rock dissolution and redistribution, the 613C value of the precipitating mineral would be close to the average value of the

Controls on accretion of flysch and mélange belts at convergent margins: Evidence from the Chugach Bay thrust and Iceworm mélange, Chugach accretionary wedge, Alaska

Controls on accretion of flysch and mélange terranes at convergent margins are poorly understood. Southern Alaska's Chugach terrane forms the outboard accretionary margin of the Wrangellia composite terrane, and consists of two major lithotectonic units, including Triassic‐Cretaceous mélange of the McHugh Complex and Late Cretaceous flysch of the Valdez Group. The contact between the McHugh Complex and the Valdez Group on the Kenai Peninsula is a tectonic boundary between chaotically deformed melange of argillite, chert, greenstone, and graywacke of the McHugh Complex and a less chaotically deformed mélange of argillite and graywacke of the Valdez Group. We assign the latter to a new, informal unit of formational rank, the Iceworm mélange, and interpret it as a contractional fault zone (Chugach Bay thrust) along which the Valdez Group was emplaced beneath the McHugh Complex. The McHugh Complex had already been deformed and metamorphosed to prehnite‐pumpellyite facies prior to formation of the Iceworm mélange. The Chugach Bay thrust formed between 75 and 55 Ma, as shown by Campanian‐Maastrichtian depositional ages of the Valdez Group, and fault‐related fabrics in the Iceworm mélange that are cut by Paleocene dikes. Motion along the Chugach Bay thrust thus followed Middle to Late Cretaceous collision (circa 90–100 Ma) of the Wrangellia composite terrane with North America. Collision related uplift and erosion of mountains in British Columbia formed a submarine fan on the Farallon plate, and we suggest that attempted subduction of this fan dramatically changed the subduction/accretion style within the Chugach accretionary wedge. We propose a model in which subduction of thinly sedimented plates concentrates shear strains in a narrow zone, generating mélanges like the McHugh in accretionary complexes. Subduction of thickly sedimented plates allows wider distribution of shear strains to accommodate plate convergence, generating a more coherent accretionary style including the fold‐thrust structures that dominate the outcrop pattern in the Valdez belt. Rapid underplating and frontal accretion of the Valdez Group caused a critical taper adjustment of the accretionary wedge, including exhumation of the metamorphosed McHugh Complex, and its emplacement over the Valdez Group. The Iceworm mélange formed in a zone of focused fluid flow at the boundary between the McHugh Complex and Valdez Group during this critical taper adjustment of the wedge to these changing boundary conditions.