Australia: The Land Where Time Began

A biography of the Australian continent 

Oceanic Crust - Origin

According to the model of petrologic processes that occur at ocean ridges that was proposed, and Widely accepted, (Cann, 1970, 1974), hot asthenospheric material rises buoyantly (Nichola et al., 1994) at a rapid enough rate up a narrow zone, passing through the basalt melting curve, providing an interstitial melt of basaltic composition. As the asthenophere rises, the volume of the molten fraction increases, eventually ascending independently of the parental material, a magma chamber being formed in the lower part of the crust at the layer 3 level. Pillowed lava flows are produced when part of the magma rises through crustal fissures, erupting on the floor of the ocean. Layer 2 of the oceanic crust is comprised of lavas and dikes. Solidification of magma in the fissures that feed the flows forms a zone of dikes beneath the flows.Extrusion and intrusion have been modelled and compared with complexes of ophiolite (Kidd, 1977). Sheeted dikes have been found to be the only component of layer 2C, the dikes having been intruded through zones less than 50 m wide. About 10 % more chilled margins is found on one side of the dikes than on the other, indicating the later dikes cut the dikes, the magma of the original dyke being found on opposites sides of the crest of the ridge. Because dike intrusion would occur preferentially where existing dikes are at their weakest, in the central axis, the hottest part of the system, this explains the symmetry of sea floor spreading about the axis of the ridge. The lavas above the dikes have been suggested to cool rapidly in contact in seawater, solidifying after flowing for less than 2 km from the ridge. It has been predicted that lavas and dikes rotate towards the crest of the ridge as they move away from the extrusion zone, a result of isostatic adjustment (Fig. 6.17, Source 1).As they equilibrate at high temperatures in the presence of seawater, they undergo metamorphism near the ridge axis (Kearey et al., 2009).

Conformation of the predictions of the Kidd (1977) model for the origin of layer 2 has come from studies of sections through the upper crust, that were revealed by major fault scarps and DSDP/ODP drill cores at hole 504B, all fast spreading Pacific crust (Karson, 2002; Section 6.9). Another prediction of the model is that the extrusive layer should be very thin and the dikes correspondingly closer to the sea floor (Fig. 6.17, Source 1), beneath the axial high. Seismic studies have revealed high seismic velocities in a narrow central band beneath the axial high (Toomey et al., 1990; Caress et al., 1992), and a thin extrusive layer that rapidly thickens away from the axis, within 1-2 km (Detrick et al., 1993b; Kent et al., 1994). This prediction has also been confirmed.

According to the model of Cann (1974), at lower levels, the crust develops from crystallisation of the axial magma chamber. Olivine and chrome spinel are the first minerals to crystallise in the magma chamber, dropping through the magma to form a basal layer of dunite, chromite also forming occasional accumulations. As cooling progresses, pyroxene crystallises and layers of cumulate peridotitic (of olivine and pyroxene) are produced, further upwards being replaced by pyroxenites as the crystallisation begins to be dominated by pyroxene. Plagioclase crystallises ultimately and there is formation of olivine gabbros. At this point there is still a large volume of liquid that solidifies over a very small range of temperatures, forming an upper, "isotropic" gabbro. The last fraction to crystallise, essentially plagioclase and quartz, a small volatile-rich residuum of the differentiation process, sometimes forming veins and small pockets of "plagiogranite" within the overlying sheeted complex dike as it intrudes upwards. It has been suggested that in the uppermost part of the magma chamber, the abundance of volatiles, especially water, may result, at least in part, from interaction with seawater that percolates down, and/or the stoping of the overlying dikes, that have been hydrothermally altered, into the magma chamber.

Seismic layers 3A and 3B often correlate with the 2 gabbro units, isotropic and layered respectively. The sub-Moho seismic velocities would then be accounted for by the olivene- and pyroxene-rich ultramafic cumulates, the Moho occurring at the base  of the mafic section within the crystallised magna chamber. The uppermost ultramafics may become partially hydrated (serpentinized) in the lower temperature environment off axis, acquiring lower seismic velocities that are more characteristic of layer 3B as a result. The seismic Moho would then occur within the ultramafic section at greater depth. Because of this uncertainty, petrologists have had a tendency to define the base of the crust as the base of the presumed magna chamber, i.e. the dunite/chromatite horizon, hence the "petrologic Moho".

The known petrology of oceanic crust formed at ridge crests that are spreading rapidly, where a steady state magma chamber exists, have been largely explained by the model of Cann (1974) and Kidd (1977). It appears probable that magma chambers may be transient at ridge crests that are spreading more slowly, and the crustal acceleration zone is wider. The alternative model derived from early reinterpretations of ophiolites in terms of sea floor spreading may possibly be more appropriate in this case. Multiple intrusive relationships have been observed at all levels in southern Cypress in the Troodos ophiolites (Moores & Vine, 1971), suggesting that in the main layer of the crust there are many small magma chambers. This model is favoured by some to explain the production of oceanic crust at ridge crests that are slow-spreading (Smith & Cann, 1993). The magma supply may be much lower away from the centres of segments, especially near transform faults. In thinned oceanic crust serpentinised peridotite from the mantle appears to be commonly present. On very slow-spreading ridges this type of crust is even more common, most of the crust being ultimately what is effectively exposed mantle, with or without a basalt covering. The crust is essentially mantle peridotite that has been serpentinised and highly tectonised on ultra-slow ridges, such as the Gakkel Ridge, with volcanic centres at intervals of 100 x 50 km.

Thermal modelling has been used as an alternative approach to studying the mid-ocean ridge accretionary processes (Sleep, 1975; Kusznir & Bott, 1976; Chen & Morgan, 1990). The inclusion in models of hydrothermal circulation at the ridge crests and the different rheological properties of the crust compared to the mantle, at high temperatures the crust is more ductile than the mantle, resulted in significant improvements (Chen & Morgan, 1990). The rate the magma is supplied to the crust influences the thermal regime beneath ridge crests, the rate of supply depending on the rate of spread. At a fast-spreading ridge, compared to one that is spreading more slowly, and has a lower rate of magma supply, the brittle-ductile transition occurs at about 750o C at a shallower depth in the crust. This implies there is a much greater volume at a fast-spreading ridge, hence width of ductile lower crust. The overlying brittle crust is effectively decoupled by the ductile crust, from the viscous drag of the convecting mantle beneath it, the tensile stresses pulling the plates apart being concentrated in a layer that is relatively thin and weak, extending by tensile fractures in the very narrow zone along the ridge axis. The brittle layer is thicker, with a much smaller volume of ductile crust, on a slow-spreading ridge. This results in the tensile stresses being distributed over a larger area, with more viscous drag on the brittle crust. The upper brittle layer deforms, under these conditions, by steady state attenuation or "necking", a large number of normal faults being created a median valley.

It has been shown that the transition from smooth topography with a buoyant axial high to a median rift valley is abrupt, for crust that is of normal thickness and appropriate model parameters, at a full spreading rate of about 70 mm/year, as observed (Chen & Morgan, 1990). For thicker crust forming at a ridge that is slow spreading, such as the Reyk-janes Ridge, immediately to the south of Iceland, the model predicts a much larger volume of ductile crust, smooth topography being developed instead of a rift valley. On a slow-spreading ridge, where the crust is thin, such as in the vicinity of fracture zones on the mid-Atlantic Ridge, the median valley will be more pronounced than at the centre of a segment. Areas of higher or lower than normal upper mantle temperatures respectively tend to be areas where the crust is thicker or thinner than normal, enhancing the effect in each case. The model has been extended to incorporate a magma chamber that was observed on the East Pacific Rise (Morgan & Chen, 1993). The extended model predicts that a steady state magma chamber can only exist where spreading rates are above 50 mm/year, the depth to the top of the chamber decreasing as the rate of spreading increases, while retaining the essential features of the Chen & Morgan (1990) model.

According to Kearey Iet al., there is generally good agreement between the models for oceanic crust formation and observations on in situ ocean floor, as well as on ophiolites, though some aspects are still problematic. A process that is not well understood is the evolution of the median valley as accretion occurs, the way the flanks are uplifted, and the ultimate reversing of the normal faults. For amagmatic segments of ridges that are spreading at rates that are very slow to ultra slow, places where mantle material is emplaced directly on the seafloor, this is of particular significance. Another highly debated process is the gabbroic layer 3, from magma chambers that are steady state or transient.

See Source 1 for more detailed information on all aspects of plate tectonics

Sources & Further reading

  1. Kearey, Philip, Klepeis, Keith A. & Vine, Frederick J., 2009, Global Tectonics, 3rd Edition, Wiley-Blackwell.
Author: M. H. Monroe
Last Updated 16/05/2011



Journey Back Through Time
Experience Australia
Aboriginal Australia
National Parks
Photo Galleries
Site Map
                                                                                           Author: M.H.Monroe  Email:     Sources & Further reading