Axial Rift Valley Research Articles

Dispersion of deep-sea hydrothermal plumes at the Endeavour Segment of the Juan de Fuca Ridge: a multiscale numerical study

A multiscale numerical framework has been developed to investigate the dispersion of deep-sea hydrothermal plumes that originate from the Endeavour Segment of the Juan de Fuca Ridge located in the Northeast Pacific. The analysis of simulation outputs presented in this study provides insights into the influences of tidal forcing and the buoyancy flux associated with hydrothermal venting on ocean circulation and plume dispersion in the presence of pronounced seafloor topography. The results indicate that tidal forcing drives anti-cyclonic circulation near the ridge-axis, while hydrothermal venting induces cyclonic circulation around vent fields within the axial rift valley. Tidal forcing has a notable impact on plume dispersion, particularly near the large topographic features to the north of the Endeavour Segment. Furthermore, plume dispersion exhibits notable inter-annual variability, with a northbound trajectory in 2016 and a southbound trajectory in 2021. The study also reveals that both buoyancy fluxes and tidal forcing enhance the mixing of hydrothermal plumes with ambient seawater.

Geophysical investigation of the Mado Megamullion oceanic core complex: implications for the end of back-arc spreading

Detachment faulting is one of the main styles of seafloor spreading at slow to intermediate mid-ocean ridges. However, we have limited insight into its role in back-arc basin formation. We surveyed a remnant back-arc spreading center in the Philippine Sea and determined the detailed features and formation processes of the Mado Megamullion (Mado MM) oceanic core complex (OCC). This was undertaken in the context of back-arc evolution, based on the shipborne bathymetry, magnetics, and gravity with radiometric age dating of the rock samples collected. The Mado MM OCC has a typical OCC morphology with prominent corrugations on the domed surface and positive gravity anomalies, suggesting that there has been an exposure of the lower crust and mantle materials by a detachment fault. The downdip side of the detachment continues to the relict axial rift valley, which has indicated that the Mado MM OCC was formed at the end of the back-arc basin opening. The spreading rate of the basin decreased once when the spreading direction changed after six million years of stable trench perpendicular spreading. The rate then further decreased immediately prior to the end of the spreading when the Mado MM OCC was formed. The existence of other OCC-like structures in the neighboring segment and the previously reported OCCs along the Parece Vela Rift have indicated that the melt-poor, tectonic-dominant spreading is a widespread phenomenon at the terminal phase of back-arc spreading. The decrease in spreading rate in the later stage is consistent with the previous numerical modeling because of the decrease in trench retreat. In the Izu–Bonin–Mariana arc trench system, the rotation of the spreading axis and the resultant axis segmentation have enhanced the lithosphere cooling and constrained mantle upwelling, which caused the tectonic-dominant spreading at the final phase of the basin evolution.

Reverse-faulting earthquakes and the tectonics of slowly-spreading mid-ocean ridge axes

We examine the distribution and characteristics of 70 reverse-faulting earthquakes at slowly-spreading ridges in the Atlantic and Indian Oceans. All were large enough (Mw≥4.7) for their focal mechanisms to be determined and listed in the gCMT catalogue (www.globalcmt.org/). Of these, 21 were sufficiently large (Mw≥5.3) for detailed body-waveform analysis to determine accurate mechanisms, centroid depths and water depths at their epicentres. All had shallow centroids 1–3 km below the sea bed, and occurred beneath gravity and bathymetric highs on ridge flanks, not in the axial valley, consistent with the locations of reverse-faulting microearthquakes discovered in earlier OBS experiments. We describe a simple quantitative model that explains their occurrence by the bending of an elastic layer about 3 km thick as it is uplifted and exhumed into the dome of a core complex by detachment faulting. The widespread distribution of such events, and their global seismic moment-release rate, suggests that asymmetric spreading in core complexes is the dominant mode of extension at such slowly-spreading ridges. Such an inherently asymmetric process is capable of producing a steady-state axial rift valley, and the overall symmetry of plate creation, but only if its polarity reverses in space and time along the axis of the ridge, for which there is some evidence in the gravity from altimetry.

Tectonic termination of oceanic detachment faults, with constraints on tectonic uplift and mass wasting related erosion rates

New high-resolution bathymetric data from Atlantis Massif and surrounding seafloor (30°N, Mid-Atlantic Ridge) records avolcanic extension associated with the formation of the axial rift valley floor, following the tectonic truncation of an active corrugated oceanic detachment fault system. The truncated Atlantis detachment is tectonically uplifted by a high-angle valley-bounding normal fault, formed after westward migration of the ridge at ∼0.4 to 0.1 Ma. Detachment fault remnants, with preserved corrugations, lie within the present-day rift valley seafloor, and demonstrate that a ∼20 km ridge section in the immediate vicinity of the Atlantis Fracture zone has not recorded any recent volcanic activity. Avolcanic extension may thus occur locally at the slow-spreading Mid-Atlantic Ridge, albeit for limited periods of time (less than a few hundred thousand years). The new fault dissecting the detachment shows a throw of ∼2800 m, partly due to flexural uplift. Emplacement of the Lost City hydrothermal site occurred at a late stage post-dating the detachment truncation and avolcanic rift valley formation. From the inferred timing of the westward ridge axis shift we calculate uplift rates ≥ 7 mm/yr, possibly as high as 33 mm/yr, which are equivalent to or greater than the fastest vertical uplift rates of active normal faults measured to date on Earth (Gulf of Corinth). Geomorphologic observations also demonstrate that mass wasting efficiently reworks the seafloor topography. We obtain local incision and erosion rates ≥1-2 mm/yr locally, and as high as 4-8 mm/yr, depending on the assumed age for the rift bonding fault (0.4 vs. 0.1 Ma respectively). Our results suggest that (1) avolcanic extension may occur locally at the slow-spreading Mid-Atlantic Ridge, albeit for limited periods of time (less than a few 100s of kyrs), and (2) document that shifts in axial valley location related to the abrupt abandonment of detachment faults is a first-order process in the asymmetric accretion of slow-spread oceanic lithosphere.

Базальты Быстринской зоны из источников континентальной литосферной мантии: Тункинская долина Байкальской рифтовой системы

In terms of petrochemical and trace-element data, basalts from the axial rift valley are suggested to originate from sources of the sub-continental lithospheric mantle, inhomogeneity of which was due to extracting upper and lower crustal components. Trace-element modeling indicated 3-8 %partial melting of apatite-bearing sources with varying garnet and clinopyroxene proportions.

Distinctive Seafloor Fabric Produced Near Western Versus Eastern Ridge‐Transform Intersections of the Northern Mid‐Atlantic Ridge: Possible Influence of Ridge Migration

AbstractMultibeam bathymetry compiled along fracture zones of the northern Atlantic reveals a striking morphological asymmetry. Seafloor fabric produced at western ridge‐transform intersections (RTIs) tends to consist of linear ridges, small in amplitude, and regularly spaced, and oceanic core complexes (OCCs) occur infrequently. In contrast, seafloor fabric produced at eastern RTIs is more irregular and blocky and displays characteristics usually associated with melt‐poor accretion: These include more than double the occurrence of OCCs as well as greater seafloor depths, even where seafloor is younger than that on the opposite side of the fracture zone. We propose that this asymmetry is a consequence of the westward migration of the Mid‐Atlantic Ridge. Such migration is expected to result in an enhanced melt supply at leading (western) RTIs compared to trailing (eastern) RTIs. The morphological asymmetry is observed for ridge offsets of ~40 to ~200 km, a range that may be related to the width of melting regimes that supply ridge segments. At slow spreading ridges, contrasting melt supplies across fracture zones may therefore be best expressed in the distinct seafloor fabrics preserved beyond the dynamically maintained relief of the axial rift valley and walls rather than in the contrasting axial depths documented for faster spreading ridges. Although OCCs appear to form preferentially at spreading rates lower than 30 mm/year, their common occurrence along the northern Mid‐Atlantic Ridge may also reflect the significantly higher rates at which it is migrating compared to the southern Mid‐Atlantic Ridge.

Microearthquake activity, lithospheric structure, and deformation modes at an amagmatic ultraslow spreading Southwest Indian Ridge segment

AbstractWhile nascent oceanic lithosphere at slow to fast spreading mid‐ocean ridges (MOR) is relatively well studied, much less is known about the lithospheric structure and properties at ultraslow MORs. Here we present microearthquake data from a 1 year ocean bottom seismometer deployment at the amagmatic, oblique supersegment of the ultraslow spreading Southwest Indian Ridge. A refraction seismic experiment was performed to constrain upper lithosphere P‐velocities and results were used to construct a 1D velocity model for earthquake location. Earthquake foci were located individually and subsequently relocated relative to each other to sharpen the image of seismically active structures. Frequent earthquake activity extends to 31 km beneath the seafloor, indicating an exceptionally thick brittle lithosphere and an undulating brittle‐ductile transition that implies significant variations in the along‐axis thermal structure of the lithosphere. We observe a strong relation between petrology, microseismicity distribution, and topography along the ridge axis: Peridotite‐dominated areas associate with deepest hypocenters, vast volumes of lithosphere that deforms aseismically as a consequence of alteration, and the deepest axial rift valley. Areas of basalt exposure correspond to shallower hypocenters, shallower and more rugged axial seafloor. Focal mechanisms deviate from pure extension and are spatially variable. Earthquakes form an undulating band of background seismicity and do not delineate discrete detachment faults as common on slow spreading ridges. Instead, the seismicity band sharply terminates to the south, immediately beneath the rift boundary. Considering the deep alteration, large steep boundary faults might be present but are entirely aseismic.

Ilhas oceânicas brasileiras e suas relações com a tectônica atlântica

Sintetizam-se conhecimentos modernos sobre as ilhas e arquipélagos oceânicos brasileiros, cuja origem e localização geográfica sofreu decisiva influência de estruturas da litosfera oceânica. Reúnem-se informações sobre a geologia das ilhas e a nomenclatura recente, química e mineralógica, de suas rochas. É sugestiva a influência de hotspots na evolução de zonas de fratura, originando cadeias vulcânicas e ilhas. No Ceará a atuação de hotspot suposto astenosférico, na extensão ao continente de zona de fratura reativada, desenvolveu estruturas, magmatismo alcalino, alto fluxo térmico e sismicidade. Penetrando nas fraturas reativadas dessa zona no oceano, com o deslizamento da placa formou-se uma cadeia vulcânica, em cuja extremidade está o Arquipélago de Fernando de Noronha. Essa evolução é similar à da Ilha de Trindade e Arquipélago Martin Vaz, que constituem o topo emerso de grandes montes vulcânicos do extremo oriental das duas cadeias, situadas em zona de fratura que penetra na borda do continente. O estranho Aarquipélago de São Pedro e São Paulo possui origem inteiramente diversa, pois emerge no trecho transformante ativo de uma cadeia de zona de fratura onde esta secciona o rift-valley axial da Dorsal Meso-Atlântica, expondo-se uma protrusão de rochas mantélicas sem vulcânicas emersas aflorantes, embora as existam submersas.

Late Pleistocene cyclicity of sedimentation and spreading-center structure in the Central Gulf of California

Abstract The interaction between climatic-driven fluctuations in sediment supply and dynamic changes in basin physiography is a fundamental process of rift basin evolution that is poorly understood. A high-resolution seismic profile collected across the southern Guaymas Basin spreading center reveals how cyclical changes in sedimentation interact with on-going axial rifting and accretion of oceanic crust. The 4 km-wide axial rift valley abuts the steep, canyon incised continental slope of Baja California. Alternating acoustically transparent and well-layered high-amplitude seismic units imaged on the accreted flanks record the filling and regeneration of the axial rift over the past ~ 200 ka. An approximate correlation of seismic units to nearby Deep Sea Drilling Program (DSDP) Site 478 indicates that transparent units are composed primarily of diatomaceous silty mud turbidites, whereas the intervening well-layered high-amplitude units represent mud turbidites interlayed with terrigenous silty sand. Age estimates using sparse biostratigraphy and plate stratigraphy suggest deposition of transparent units coincident with interglacial/high-stands, while deposition of well-layered high-amplitude units corresponds to glacial/low-stands. Analysis of seismic stratigraphy and imaged faults reveals that during glacial/low-stands, deposition of terrigenous-rich turbidites out-paced axial spreading and subsidence, filling the axial rift and spilling over onto the adjacent basin floor. However, during interglacial/high-stands, reduction in sediment supply resulted in reestablishment of the axial rift relief. We propose that increases in sediment supply and terrigenous material during glacial periods were caused by a combination of (i) enhanced erosion of Baja California during pluvial periods, corresponding to northern hemisphere glacial maxima, (ii) increased delivery of sediment from the Sierra Madre Occidental, and (iii) sediment bypass on the continental shelf and slope to the deep-water basins during relative sea-level low-stands.

Hydrothermal circulation within the Endeavour Segment, Juan de Fuca Ridge

Areas of the seafloor at mid‐ocean ridges where hydrothermal vents discharge are easily recognized by the dramatic biological, physical, and chemical processes that characterize such sites. Locations where seawater flows into the seafloor to recharge hydrothermal cells within the crustal reservoir are by contrast almost invisible but can be indirectly identified by a systematic grid of conductive heat flow measurements. An array of conductive heat flow stations in the Endeavour axial valley of the Juan de Fuca Ridge has identified recharge zones that appear to represent a nested system of fluid circulation paths. At the scale of an axial rift valley, conductive heat flow data indicate a general cross‐valley fluid flow, where seawater enters the shallow subsurface crustal reservoir at the eastern wall of the Endeavour axial valley and undergoes a kilometer of horizontal transit beneath the valley floor, finally exiting as warm hydrothermal fluid discharge on the western valley bounding wall. Recharge zones also have been identified as located within an annular ring of very cold seafloor around the large Main Endeavour Hydrothermal Field, with seawater inflow occurring within faults that surround the fluid discharge sites. These conductive heat flow data are consistent with previous models where high‐temperature fluid circulation cells beneath large hydrothermal vent fields may be composed of narrow vertical cylinders. Subsurface fluid circulation on the Endeavour Segment occurs at various crustal depths in three distinct modes: (1) general east to west flow across the entire valley floor, (2) in narrow cylinders that penetrate deeply to high‐temperature heat sources, and (3) supplying low‐temperature diffuse vents where seawater is entrained into the shallow uppermost crust by the adjacent high‐temperature cylindrical systems. The systematic array of conductive heat flow measurements over the axial valley floor averaged ∼150 mW/m2, suggesting that only about 3% of the total energy flux of ocean crustal formation is removed by conductive heat transfer, with the remainder being dissipated to overlying seawater by fluid advection.

Subsidence and crustal roughness of ultra-slow spreading ridges in the northern North Atlantic and the Arctic Ocean

SUMMARY Five basin-wide seismic reflection profiles of up to 550 km each were acquired in the Arctic Ocean and the northern North Atlantic in 2001 and 2002. The main objective was to investigate the depth to the basement and to analyse the crustal structure, morphology and roughness of ultra-slow spreading ridges of the Gakkel, Molloy and Knipovich ridges. To date, little is known to date of the ultra-slow spectrum of such spreading ridges. The seismic profiles of all investigated ridges show similar morphological characteristics with deep axial valleys and rough basement topography. Magnetic data compilation and interpretation suggests that the ultra-slow spreading systems are fairly stable and existed during the entire evolution of the basins to the north of the Greenland Fracture Zone. The thermal subsidence curve was calculated and corrected for sediment loads, and crustal roughness values are estimated for all five profiles. The resulting roughness values append the global roughness data set for ultra-slow spreading systems. The results are higher than those predicted by interpolating existing global roughness. This study confirms the presence of a global relationship between crustal roughness, ridge morphology and spreading rates. New curve fits, supporting the global relationship, are discussed. Data on present spreading rates, ridge morphology, subsidence and roughness provide a better insight into the development of the axial ridge morphology in the study area. The results show that the basins to the north of the Greenland Fracture Zone were formed at ultra-slow spreading axial rift valleys and continued spreading at ultra-slow rates to the present day configuration.

Hydrothermal alteration and sulfide mineralization in gabbroids of the Markov Deep (Mid-Atlantic Ridge, 6° N)

A new type of sulfide occurrence related to metasomatically altered brecciated gabbroids has been studied at the Sierra Leone site situated in the axial rift valley of the Mid-Atlantic Ridge (Markov Deep, 6° N). Two associations of plutonic, subvolcanic, and volcanic rocks were dredged: (1) mid-ocean ridge basalts (MORB) and their intrusive analogues and (2) rocks of the silicic Fe-Ti-oxide series with dominating gabbronorites and sporadic trondhjemites. Almost all igneous rocks at the Sierra Leone site are enriched in Pb, Cu, U, Ga, Ta, Nb, Cs, and Rb and depleted in Zr, Th, and Hf. The rocks of the Fe-Ti-oxide series are enriched in Zn, Sn, and Mo and depleted in Ni and Cr. The main ore-bearing zone is situated at the foot of the eastern wall of the deep, where it is hosted in cataclastic hornblende gabbro and gabbronorite of the Fe-Ti-oxide series. Ore mineralization in metasomatically altered rocks is composed of quartz-sulfide and prehnite-sulfide veinlets, disseminated sulfide, and veined copper sulfide ore. The ore consists of pyrite, chalcopyrite, sphalerite, pyrrhotite, bornite, chalcocite, and digenite. The δ34S value of sulfides varies from 3.0 to 15.3‰. At the foot of the eastern wall of the Markov Deep, directly downslope from the ore-bearing zone, loose sediments contain grains of native Cu, Pb, Zn, and Sn and intermetallic compounds (isoferroplatinum, tetraferroplatinum, and brass) apparently derived from the ore. Mineral assemblages of ore-bearing metasomatic rocks and fluid inclusions therein were studied. Ore metasomatism developed under a low oxygen potential within a temperature interval from 400 to 160°C, though initial hydrothermal alteration of rocks proceeded at temperatures of 800–450°C. The temperature of stringer-disseminated ore mineralization is estimated at 170–280°C. The hydrothermal fluids are considered to be of magmatic origin; as the hydrothermal system evolved, they became diluted with seawater that was contained in fractured oceanic crust. The ore matter could have been derived from magmatic fluids that were released from water-saturated melts of the Fe-Ti oxide series during their ascent and leached from host gabbroids in the process of metasomatic alteration.

Topographic controls on dike injection in volcanic rift zones

Dike emplacement in volcanic rift zones is often associated with the injection of “blade-like” dikes, which propagate long distances parallel to the rift, but frequently remain trapped at depth and erupt only near the tip of the dike. Over geologic time, this style of dike injection implies that a greater percentage of extension is accommodated by magma accretion at depth than near the surface. In this study, we investigate the evolution of faulting, topography and stress state in volcanic rift zones using a kinematic model for dike injection in an extending 2-D elastic-viscoplastic layer. We show that the intrusion of blade-like dikes focuses deformation at the rift axis, leading to the formation of an axial rift valley. However, flexure associated with the development of the rift topography generates compression at the base of the plate. If the magnitude of these deviatoric compressive stresses exceeds the deviatoric tensile stress associated with far-field extension, further dike injection will be inhibited. In general, this transition from tensile to compressive deviatoric stresses occurs when the rate of accretion in the lower crust is greater than 50–60% of the far-field extension rate. These results indicate that, over geologic timescales, the injection of blade-like dikes is a self-limiting process in which dike-generated faulting and topography result in an efficient feedback mechanism that controls the time-averaged distribution of magma accretion within the crust.

Ancient melt extraction from the oceanic upper mantle revealed by Re–Os isotopes in abyssal peridotites from the Mid-Atlantic ridge

Comprehensive major-, trace-element and rhenium–osmium (Re–Os) isotope data are presented for abyssal peridotites from Ocean Drilling Program (ODP) Leg 209, in the North Atlantic. The samples are from a single core (Site 1274A) located on the western wall of the axial rift valley of the Mid-Atlantic ridge, and their study allows elemental and isotope information to be precisely related to spatial variations in primary lithology, serpentinisation and seafloor weathering. The harzburgites and dunites at this site are highly serpentinised (with the degree of serpentinisation increasing with depth below the sea floor). Petrographic observations and variations of fluid mobile elements (such as Ba, Sr, U, and Re) are consistent with seawater interaction in the upper part of the core. Nevertheless, major and trace element indicators of the extent of melt depletion indicate extreme melt loss, and suggest that these are amongst the most depleted abyssal peridotites recovered thus far. Despite the evidence for extensive serpentinisation and sea floor weathering all of the samples possess 187Os/ 188Os isotope compositions that are less radiogenic than estimates for the primitive upper mantle (PUM), lower than any yet reported for abyssal peridotites, and consistent with melt loss over, at least, the past 1.5 Ga. Single sulphide Re–Os data show evidence for recent recrystallisation or diffusional modification either due to partial melting or seawater alteration. However, some grains are extremely unradiogenic ( 187Os/ 188Os = 0.114) providing unequivocal evidence for at least some degree of melt depletion at ca. 2 Ga. Taken with recently published data these results suggest that ancient melt depletion may be a widespread feature of the oceanic upper lithosphere, even though evidence for this depleted reservoir has not yet been observed in the Re–Os chemistry of mid-ocean ridge basalts.

Magmatic segmentation of the northern Knipovich Ridge: Evidence for high‐pressure fractionation at an ultraslow spreading ridge

The ultraslow spreading Knipovich Ridge shows axial topographic highs that are associated with off‐axis linear arrays of seamounts that parallel the spreading direction. These linear arrays suggest that the segmentation has been stable for at least 7–8 m.y. Axial topographic highs are marked with distinct gravity anomalies, suggesting that they represent thicker crust and hence volcanic segment centers. Dredging on the northern Knipovich Ridge between 76°30′ and 77°50′N shows that the axial and off‐axial seamounts, and the deeper axial rift valley between the segment centers, are composed of volcanic rocks. The basalts are subdivided into high‐, intermediate‐, and low‐FeO groups. Low‐FeO basalts, with generally more enriched radiogenic isotope ratios and REE patterns, were only dredged midway between segment centers, whereas dredges close to the segment centers comprised high‐ and intermediate‐FeO basalts. The variations in isotopes and REE patterns suggest that the mantle below the northern Knipovich Ridge is heterogeneous, and the along‐axis variations can be explained by higher degrees of melting of this heterogeneous source close to segment centers. The high‐ and intermediate‐FeO basalts define different MgO‐FeO liquid lines of descents that may reflect differences in the crystallization pressures or the water contents of the magmas. Similar water contents of the high‐ and intermediate‐FeO basalts suggest that the differences in FeO primarily are a pressure effect and that the intermediate‐FeO group fractionated at higher pressure than the high‐FeO group. The high‐FeO basalts show both low‐ and high‐pressure phenocryst assemblages, suggesting crystallization at high pressures (6–7 kbar) followed by shallow, low‐pressure fractionation. On the basis of the new data we propose that magmas at this ultraslow spreading ridge fractionate at mantle depth, except at segment centers where magma productivity occasionally is high enough to sustain shallow level magma reservoirs.

Autonomous hydrophone array monitors seismic activity at northern Mid‐Atlantic Ridge

The northern Mid‐Atlantic Ridge (MAR) is a classic, slow‐spreading ridge with an average spreading rate of ∼25 mm/yr. The MAR is typically marked by a 1.5–3‐km‐deep, 15–30‐km‐wide axial rift valley Major transforms offset the ridge along its length. Between the transforms, the ridge is divided into spreading segments; their ends are defined by non‐transform offsets. The rift valley commonly contains an axial volcanic ridge that is considered to be the predominant site of volcanic activity [e.g., Ballard and van Andel, 1977]. Based on the sizes of the volcanic ridges [e.g., Smith and Cann, 1999], large volumes of lava must be erupted on the valley floor. Currently, however, it is not known how frequently magmatic events occur, whether they migrate along the axis, whether there are spatial patterns in these events, or how they vary through time.

Evidence for a plate tectonics debate

A central problem in understanding the Earth system is the relationship between mantle convection and near-surface structure, geophysics and geochemistry. Unexplained anomalies of the plate tectonics (PT) model are the postulated mechanism of generating the axial rift valley by upward movement of marginal fault blocks, contrary to observations in the Icelandic rift, and the mechanism by which a postulated broad upwelling plume can yield the required narrow zone of axial volcanism. These discrepancies are the basis for a critical reappraisal of other evidence. It is shown that the generation of oceanic magnetic stripes, which led to the ingenious spreading hypothesis, is a result of narrowing of a formerly expanded ocean ridge volcanic system, and resultant sequential cooling of crust and upper mantle, and does not require ocean floor spreading. An alternative ridge model postulates sub-continent upwelling, sub-ridge convergent flow and generation of mid-ocean ridge basalt (MORB) by heterogeneous volatile-promoted melting at sub-axial zones where oceanic crust is recycled into the mantle. That model is favored by the results of convection experiments and confirmed by a series of independent indicators: heat flow beneath ridge flanks, coriolis curvature of fracture zones (FZ), downstream development of convective rolls and near-transform tectonic rotations. Sub-axial downflow is confirmed by North Atlantic positive geoid anomalies, by high P-wave velocities deep below the ridge axis, and by the synclinal “flexload” structure and compressional stress regime of near-axial crust. Oceanic island volcanism is attributed to the same process, crustal recycling at local sites of downflow focused by deep residual masses that are relatively cold (radioactivity depleted) and viscous (volatile depleted). The recycling model is confirmed, for both MORB and OIB, and decompression melting of mantle plumes rejected, on several grounds: (1) local recycling of crust and resultant mantle hydration are indicated by diapirs of serpentinized mantle along fracture zones and ridge axes; (2) distinctive isotopic signatures in oceanic basalt require local recycling of crust and sediments to the magma source; (3) the principal oceanic island groups, exemplified by the Hawaiian islands, lack the positive heat flow anomaly expected over a plume and are underlain by zones of relatively high seismic velocities. The proposed local recycling accounts for several previously puzzling features of global geochemical systems, including the carbon budget, the lead paradox and the isotopic array of heavy noble gases in oceanic basalt. Deduced end-member components, defined by isotope ratios, can be attributed to regionally variable mixtures of upper mantle with terrigenous and pelagic sediments, subaerial and suboceanic basalt, and with depleted lower mantle, proposed to have been available during a Pacific-centered Mesozoic mantle surge. There is no need to appeal to the several currently favored isolated reservoirs in the upper and lower mantle. Seismic data for continental ranges, exemplified by the Alpine region, show that the plate collision model is misleading in view of increasing evidence for upper mantle deformation by viscous creep. Continental rift systems are re-examined on the basis that the oceanic axial zone can be traced to the African rifts, via Afar, and are therefore taken to be a result of a similar dynamic system. It is shown that dominant upwelling plumes are focused beneath the thick keels of ancient cratons and that rift systems, typically at craton margins, are collapse structures in zones of convergence and downflow.

The role of the Spitsbergen shear zone in determining morphology, segmentation and evolution of the Knipovich Ridge

In 1989–1990 the SeaMARC II side-looking sonar and swath bathymetric system imaged more than 80 000 km2 of the seafloor in the Norwegian-Greenland Sea and southern Arctic Ocean. One of our main goals was to investigate the morphotectonic evolution of the ultra-slow spreading Knipovich Ridge from its oblique (115° ) intersection with the Mohns Ridge in the south to its boundary with the Molloy Transform Fault in the north, and to determine whether or not the ancient Spitsbergen Shear Zone continued to play any involvement in the rise axis evolution and segmentation. Structural evidence for ongoing northward rift propagation of the Mohns Ridge into the ancient Spitsbergen Shear Zone (forming the Knipovich Ridge in the process) includes ancient deactivated and migrated transforms, subtle V-shaped-oriented flank faults which have their apex at the present day Molloy Transform, and rift related faults that extend north of the present Molloy Transform Fault. The Knipovich Ridge is segmented into distinct elongate basins; the bathymetric inverse of the very-slow spreading Reykjanes Ridge to the south. Three major fault directions are detected: the N-S oriented rift walls, the highly oblique en-echelon faults, which reside in the rift valley, and the structures, defining the orientation of many of the axial highs, which are oblique to both the rift walls and the faults in the axial rift valley. The segmentation of this slow spreading center is dominated by quasi stationary, focused magma centers creating (axial highs) located between long oblique rift basins. Present day segment discontinuities on the Knipovich Ridge are aligned along highly oblique, probably strike-slip faults, which could have been created in response to rotating shear couples within zones of transtension across the multiple faults of the Spitsbergen Shear Zone. Fault interaction between major strike slip shears may have lead to the formation of en-echelon pull apart basins. The curved stress trajectories create arcuate faults and subsiding elongate basins while focusing most of the volcanism through the boundary faults. As a result, the Knipovich Ridge is characterized by Underlapping magma centers, with long oblique rifts. This style of basin-dominated segmentation probably evolved in a simple shear detachment fault environment which led to the extreme morphotectonic and geophysical asymmetries across the rise axis. The influence of the Spitsbergen Shear Zone on the evolution of the Knipovich Ridge is the primary reason that the segment discontinuities are predominantly volcanic. Fault orientation data suggest that different extension directions along the Knipovich Ridge and Mohns Ridge (280° vs. 330°, respectively) cause the crust on the western side of the intersection of these two ridges to buckle and uplift via compression as is evidenced by the uplifted western wall province and the large 60 mGal free air gravity anomalies in this area. In addition, the structural data suggest that the northwards propagation of the spreading center is ongoing and that a `normal' pure shear spreading regime has not evolved along this ridge.

Gravity evidence of very thin crust at the Gakkel Ridge (Arctic Ocean)

Gakkel Ridge, the active spreading center in the Arctic Ocean, is the slowest spreading portion of the global mid-ocean ridge system. Total spreading rates range from 0.6 cm/yr in the east where the ridge disappears beneath the Laptev shelf to 1.3 cm/yr in the west near Greenland. Bathymetry and gravity surveys of four sections of the Gakkel Ridge were carried out in 1996 by the U.S. Navy nuclear submarine USS POGY as part of SCICEX 96 in order to sample variations in seafloor morphology and gravity anomalies as a function of spreading rate. The ridge axis throughout the survey area is characterized by a continuous axial rift valley similar to that observed at other slow spreading ridges. The continuous rift axis suggests that well-organized seafloor spreading is occurring at total spreading rates of less than 1 cm/yr. In three faster spreading (1.13–1.24 cm/yr) western survey areas located between 7°E and 54°E, the Gakkel Ridge is deep compared with other ridge axes. Axial depths range between 4600 and 5100 m and ridge flanks at about 3200 m. The ridge flank morphology is very blocky and is characterized by large scarps and deep fault-bounded troughs. Very large amplitude free-water anomalies with peak-to-trough amplitudes of 85–150 mGal are observed centered on the axis of the Gakkel Ridge. Modeling of the free-water anomalies by varying the crustal thickness and average crustal density, including the gravity effect of the cooling of the mantle away from the axis, implies that if the average crustal density is less than 2900 kg/m 3, the crustal thickness must be less than 4 km. The axial rift valley at the fourth survey area, near 98°E where the total spreading rate is 0.99 cm/yr, is buried by sediments. The axis in this region is associated with a continuous 70 mGal gravity minimum implying the presence of a large buried rift valley. The rift flanks at 95°E are at a depth of greater than 3800 m, 600 m deeper than the average depth at the Gakkel Ridge axis west of 60°E. Simple isostatic calculations suggest that the crust in this region may be vanishingly thin beneath the sediment cover. These observations indicate a relationship between melt production and seafloor spreading rate at very slow spreading rates, suggesting that ultra-slow spreading may suppress melt production or delivery at the Gakkel Ridge.

Spiess Ridge: An axial high on the slow spreading Southwest Indian Ridge

We report recent mapping of Spiess Ridge with the Hawaii‐MR1 sidescan sonar. Spiess Ridge is an unusual elongate 90 by 50 km volcanic feature on the slow spreading Southwest Indian Ridge near the Bouvet hotspot. The northwest half of the ridge has a narrow ∼7‐km neovolcanic zone in MR1 sonar images and a simple magnetic anomaly sequence including possible anomalies C1n and C2n. In contrast, the southeast half shows extrusive volcanism over >40 km with distributed eruption sites, a broadened central magnetic anomaly with no anomaly C2n, and a volcanic ridge radiating from the summit of Spiess Ridge, oblique to the spreading orthogonal trend. The images show no evidence for large‐offset normal faults or an axial rift valley typical of slow spreading ridges. Overall, Spiess Ridge has an appearance very unlike that of either fast or slow spreading ridges and more transitional between a spreading ridge and a seamount. We compare the morphology of Spiess Ridge to other large volcanic structures in oceanic rifts. Spiess Ridge and Bouvet Island represent localized zones of excess melting along the Southwest Indian Ridge and we discuss their origin in the light of current ideas on ridge‐hotspot interaction.