Crustal Lid Research Articles

Models of convection and segregation in heterogeneous partially molten crustal roots with a VOF method – I: flow regimes

SUMMARYWe investigate numerically some thermomechanical conditions for the development of crustal scale diapirism and convection in a heterogeneous continental crust independently from the action of regional tectonics. Here, we consider a hot crust, with unmolten and partially molten domains of specific temperature and strain-rate dependent power-law rheologies. We take advantage of the volume of fluid (VOF) method to capture the coalescence and separation of deformable inclusions in the partially molten domain. The inclusions, of several hundred metres in size, are more or less dense and viscous with respect to the ambient medium (they also behave with a power-law rheology). We restrict our study to a 20 Myr time scale, during which gravitational dynamics may dominate over lateral tectonics and lithospheric thermal re-equilibration. The motion of these inclusions during the development of gravitational instabilities displays distinct flow regimes that depend on two Rayleigh numbers denoted RaUM and RaPM, for the unmolten and partially molten rock properties, respectively. A ‘suspension’ regime occurs at high RaUM and high RaPM, in which most of the light compositional heterogeneities remain entrained in the convective cells. In contrast at low RaUM and high RaPM, a ‘layering’ regime is characterized by merging of the light inclusions as floating clusters below the rigid upper crustal lid. This regime occurs in association with a sharp viscosity gradient at upper-to-middle crust transitional depths. In these two regimes, the dense inclusions accumulate at the bottom of the partially molten zone. Finally at moderate RaPM, a ‘diapiric’ regime reflects the segregation of the heavy and the light inclusions, respectively downward and upward, without global convection. These numerical experiments lead to a first order evaluation of the physical parameters required for the segregation of deformable inclusions of variable densities and convection, in a partially molten crust, and provide insights on the conditions for the development of migmatite domes. Geological data indicate that these processes likely occur in a large number of settings from Archean to Phanerozoic times, and contribute to the differentiation of the continental crust.

Effects of crustal thickness on magmatic differentiation in subduction zone volcanism: A global study

The majority of arc magmas are highly evolved due to differentiation within the lithosphere or crust. Some studies have suggested a relationship between crustal thickness and magmatic differentiation, but the exact nature of this relationship is unclear. Here, we examine the interplay of crustal thickness and magmatic differentiation using a global geochemical dataset compiled from active volcanic arcs and elevation as a proxy for crustal thickness. With increasing crustal thickness, average arc magma compositions become more silicic (andesitic) and enriched in incompatible elements, indicating that on average, arc magmas in thick crust are more evolved, which can be easily explained by the longer transit and cooling times of magmas traversing thick arc lithosphere and crust. As crustal thickness increases, arc magmas show higher degrees of iron depletion at a given MgO content, indicating that arc magmas saturate earlier in magnetite when traversing thick crust. This suggests that differentiation within thick crust occurs under more oxidizing conditions and that the origin of oxidation is due to intracrustal processes (contamination or recharge) or the role of thick crust in modulating melting degree in the mantle wedge. We also show that although arc magmas are on average more silicic in thick crust, the most silicic magmas (>70 wt.% SiO2) are paradoxically found in thin crust settings, where average compositions are low in silica (basaltic). We suggest that extreme residual magmas, such as those exceeding 70 wt.% SiO2, are preferentially extracted from shallow crustal magma bodies than from deep-seated magma bodies, the latter more commonly found in regions of thick crust. We suggest that this may be because the convective lifespan of crustal magma bodies is limited by conductive cooling through the overlying crustal lid and that magma bodies in thick crust cool more slowly than in thin crust. When the crust is thin, cooling is rapid, preventing residual magmas from being extracted; in the rare case that residual magmas can be extracted, they represent the very last melt fractions, which are highly silicic. When the crust is thick, cooling is slow, so intermediate melt fractions can readily segregate and erupt to the surface, where they cool and crystallize before highly silicic residual melts can be generated.

A one-dimensional seismic model for Uturuncu volcano, Bolivia, and its impact on full moment tensor inversions

Using receiver functions, Rayleigh wave phase velocity dispersion determined from ambient noise and teleseismic earthquakes, and Rayleigh wave horizontal to vertical ground motion amplitude ratios from earthquakes observed across the PLUTONS seismic array, we construct a one-dimensional (1‑D) S-wave velocity (Vs) seismic model with uncertainties for Uturuncu volcano, Bolivia, located in the central Andes and overlying the eastward-subducting Nazca plate. We find a fast upper crustal lid placed upon a low-velocity zone (LVZ) in the mid-crust. By incorporating all three types of measurements with complimentary sensitivity, we also explore the average density and Vp/Vs (ratio of P-wave to S-wave velocity) structures beneath the young silicic volcanic field. We observe slightly higher Vp/Vs and a decrease in density near the LVZ, which implies a dacitic source of the partially molten magma body. We exploit the impact of the 1-D model on full moment tensor inversion for the two largest local earthquakes recorded (both magnitude ∼3), demonstrating that the 1-D model influences the waveform fits and the estimated source type for the full moment tensor. Our 1-D model can serve as a robust starting point for future efforts to determine a three-dimensional velocity model for Uturuncu volcano.

The Precambrian supercontinent Palaeopangaea: two billion years of quasi-integrity and an appraisal of geological evidence

The Protopangaea-Palaeopangaea model for the Precambrian continental crust predicts quasi-integrity reflecting a dominant Lid Tectonics defined by a palaeomagnetic record showing prolonged near-static polar behaviour during very long time intervals (~2.7–2.2, 1.5–1.2, and 0.75–0.6 Ga). Intervening times show polar loops radiating from the geometric centre of the crust explaining the anomalous Precambrian magnetic inclination bias. The crustal lid was a symmetrical crescent-shaped body confined to a single hemisphere on the globe comparable in form to the Phanerozoic supercontinent (Neo)Pangaea. There were two major transitions in the tectonic regime when prolonged near-static motion was succeeded by widespread tectonic-magmatic activity, and each coincided with the major isotopic/geochemical signatures in the Precambrian record. The first comprised a ~90° reconfiguration of crust and mantle at ~2.2 Ga terminating the long 2.7–2.2 Ga static interval; the second was the largest continental break-up event in geological history and is constrained to the Ediacaran Period at ~0.6 Ga by multiple isotopic and geochemical signatures and the subsidence history of marine passive margins. Break-up of the lid at ~0.6 Ga defines a transition from dominant Lid Tectonics to dominant Plate Tectonics and is the primary cause of contrasts between the Precambrian and Phanerozoic aeons of geological times. The long interval of minimal continental motion in the mid-Proterozoic correlates with extensive emplacement of anorogenic anorthosite-rapakivi plutons unique to these times, and high-level emplacement was probably caused by blanketing of the mantle and comprehensive thermal weakening of the crust. Continental velocities were low during the two Proterozoic intervals characterized by profound glaciation (~2.4–2.2 and ~0.75–0.6 Ga) when partial or complete magmatic shutdown is likely to have reduced volcanic greenhouse gas production. Specific implications of Protopangaea-Palaeopangaea include: (i) support for recent evidence that 60–70% of the present continental crust had accreted by ~2.5 Ga; (ii) recognition from spatially constrained mineral provinces that sub-crustal lithosphere was already chemically differentiated by mid-Archaean times; (iii) strong axial alignment of younger greenstone belts, major Palaeoproterozoic shear zones, and later Meso–Neoproterozoic mobile/orogenic belts; (iv) concentration of anorogenic magmatism and progressive contraction of activity towards the orogenic margin subsequently to become the focus of Grenville (~1.1 Ga) orogenesis; and (v) late Neoproterozoic arc magmatism/tectonics at the instep margin of the continental crescent persisting until the Ediacaran continental break-up.

Continental velocity through Precambrian times: The link to magmatism, crustal accretion and episodes of global cooling

Quasi-integrity of continental crust between Mid-Archaean and Ediacaran times is demonstrated by conformity of palaeomagnetic poles to near-static positions between ∼2.7–2.2 Ga, ∼1.5–1.2 Ga and ∼0.75–0.6 Ga. Intervening data accord to coherent APW loops turning at “hairpins” focused near a continental-centric location. Although peripheral adjustments occurred during Early Proterozoic (∼2.2 Ga) and Grenville (∼1.1 Ga) times, the crust retained a low order symmetrical crescent-shaped form constrained to a single global hemisphere until break-up in Ediacaran times. Conformity of palaeomagnetic data to specific Eulerian parameters enables definition of a master Precambrian APW path used to estimate the root mean square velocity (vRMS) of continental crust between 2.8 and 0.6 Ga. A long interval of little polar movement between ∼2.7 and 2.2 Ga correlates with global magmatic shutdown between ∼2.45 and 2.2 Ga, whilst this interval and later slowdown at ∼0.75–0.6 Ga to velocities of <2 cm/year correlate with episodes of widespread glaciation implying that these prolonged climatic anomalies had an internal origin; the reduced input of volcanically-derived atmospheric greenhouse gases is inferred to have permitted freeze-over conditions with active ice sheets extending into equatorial latitudes as established by low magnetic inclinations in glaciogenic deposits. vRMS variations through Precambrian times correspond to the distribution of U-Pb ages in orogenic granitoids and detrital zircons and demonstrate that mobility of continental crust has been closely related to crustal tectonism and incrementation. Both periods of near-stillstand were followed by rapid vRMS recording massive heat release from beneath the continental lid at ∼2.2 and 0.6 Ga. The first coincided with the Lomagundi-Jatuli isotopic event and led to prolonged orogenesis accompanied by continental flooding and reconfiguration of the crust on the Earth's surface; the second led to continental break-up and instigated the comprehensive Plate Tectonics that has characterised Phanerozoic times. The Mesoproterozoic interval characterised by anorogenic magmatism correlates with low vRMS between ∼1.5 and 1.1 Ga. Insulation of the sub-continental mantle evidently permitted high temperature melting and weakening of the crustal lid to enable buoyant emplacement of large plutons at high crustal levels during this magmatic event unique to Mesoproterozoic and early Neoproterozoic times.

Flow of partially molten crust and origin of detachments during collapse of the Cordilleran Orogen

Abstract In metamorphic core complexes two types of detachments develop, coupled by flow of partially molten crust: a channel detachment and a rolling-hinge detachment. The channel detachment, on the hinterland side of the orogen, represents the long-lived interface that separates the partially molten crust flowing in a channel from the rigid upper crustal lid. On the foreland side of the core complex, a rolling-hinge detachment develops. This detachment dips toward the foreland, probably affects the whole crust, and its geometry is governed by strain localization at the critical interface between cold foreland and hot hinterland. Activation of the rolling-hinge detachment drives rapid decompression and melting, leading to the diapiric rise of migmatite domes in the footwall of the detachment. A kinematic hinge (switch in sense of shear) separates the two types of detachments. Structural, metamorphic and geo/thermochronological studies in the Shuswap core complex (North American Cordillera), combined with an anisotropy of magnetic susceptibility study of leucogranites concentrated in the detachments, suggest that this orogen collapsed rapidly through the development of channel and rolling-hinge detachments in the early Eocene. The kinematic hinge is currently located approximately 40 km west of the footwall in which it originated, corresponding to a mean exhumation rate of &gt;5 km Ma −1 , which explains the near-isothermal decompression recorded within the migmatite dome.

Distribution of magma beneath the East Pacific Rise near the 9°03′N overlapping spreading center from forward modeling of common depth point data

We have reprocessed six cross‐axis and three along‐axis common depth point (CDP) profiles near the 9°03′N overlapping spreading center (OSC) to understand the relationship between axial magma chamber (AMC) width and seafloor morphology. Travel time modeling of the AMC reflector reveals an asymmetric distribution of melt across the 9°03′N OSC. The variation of modeled AMC width beneath either OSC limb is minimal, but the width increases nearly fourfold across the offset attaining an estimated maximum width of 4.15 km near the 9°17′N ridge axis discontinuity. Additionally, melt distribution underlying the eastern rise limb is not symmetric with respect to the rise axis/neovolcanic zone but is displaced toward the western rise flank. Depth migration, based on a continuum velocity model consistent with postcritical reflections from the base of layer 2A, places the skewed AMC reflector beneath a nearly constant thickness sheeted dike section which dips ∼ 10° away from the rise axis. To confirm AMC continuity beneath the western rise flank, we use the Maslov synthetic seismogram method to show that amplitude enhancement of the AMC reflector is consistent with a continuous melt body underlying a thickening extrusive layer. Analysis of along‐strike CDP profiles indicates an AMC which is neither overlapping nor discontinuous when projected onto the along‐strike plane. Identifying intracrustal events on along‐axis CDP lines, however, requires extreme caution; we have modeled out‐of‐plane scattering using a Kirchhoff formulation, and we show that a coherent event identified beneath the overlap basin results from diffraction off the AMC which lies nearly 3 km to the west of the profile. We attribute the asymmetric pattern of melt to a decoupling of melt supply from preexisting weaknesses in the brittle upper crust. In this model, melt ascends upward (buoyancy forces) until deflected by the impermeable sheeted dike complex; melt then migrates updip, beneath the base of the sheeted dikes, toward the neovplcanic zone where fissuring produces a temporary conduit for emplacement. Discrete jumps in modeled AMC width toward the overlap basin represent a further displacement/defocusing of melt supply (western AMC edge) relative to the neovolcanic zone (eastern AMC edge). The asymmetric pattern of melt therefore represents a gradual, en‐echelon accommodation of melt supply across the 9 km of ridge axis offset at 9°03′N. Thus for asymmetric configurations, AMC width may not correlate solely with magmatic robustness but may signify the amount of decoupling which exists between melt supply and extrusive emplacement within the neovolcanic zone. Here we present a new model for OSC development which invokes a significant component of cross‐axis melt migration. Moreover, abrupt changes in AMC width near ridge axis discontinuities (e.g., 9°17′N deviation in axial linearity) suggest that any along‐axis melt migration is confined to subsegments of the ridge and seem to preclude the segment length migration of melt proposed in some current models. The transition of melt supply beneath the overlap basin might favor a continuous low‐velocity zone underlying this feature; if true, basin development may be related to the subsidence of a mechanically weak crustal lid. The proposed model for OSC development therefore views ridge axis discontinuities as the surficial response of misalignment and/or defocusing of melt supply in the uppermost mantle.

Lunar magma transport phenomena

An understanding of the most significant events during the first Ga or so of lunar history—global differentiation associated with a Lunar Magma Ocean (LMO) and the later generation of about 10 7 km 3 of mare basalts—depends profoundly on magma transport dynamics. LMO evolution is dominated by the characteristics of hard-turbulent convection of a fluid undergoing phase change. Quench protocrust, formed at the ocean surface in early times, is repeatedly disrupted by bolide impact, ocean body tides, and associated crustal foundering. During this epoch, the ocean temperature is set by the balance between convective heat brought to the surface and brown-body radiation from the surface. At later times, the convective flow is organized into a number of regimes. From top to bottom these include conductive crustal lid, a viscous sublayer, an inertial core, a lower viscous sublayer, and a mushy cumulate region. Crystal settling and flotation is restricted to the viscous sublayer regions. Dilatancy pumping may prevent the efficient expulsion of intercumulus melt during cumulate formation. Typical velocities within the inertial region are of order 10 m/s. Because late-stage Ti-rich cumulates are dense relative to the earlier Mg-rich cumulates, hyper-to-sub-solidus convection tends to mix LMO crystallization products. Numerical simulations that account simultaneously for compositional and thermal buoyancy effects indicate a characteristic mixing time ∼ 200 Ma. Statistical measures of mixing (variance and spatial correlation length) quantify the details of mixing. The Darcy percolative and vein-drain models of melt extraction are briefly discussed. Because compaction lengths are small, ~ 10 2 m, the rate of melt extraction is governed by the balance between melt buoyancy and Darcy friction. Differential (melt/matrix) velocities are of order 10 km/Ma for nominal values appropriate for mare basalt melt extraction from a heterogeneous (mixed) LMO source. Transport of magma through the lunar lithosphere via a connected crack network seems highly probable. For the range of inferred discharge rates applicable for Lunar Mare Volcanism (LMV; 10 3 to 10 7 m 3/s) typical fracture widths and ascent speeds are 10 −1 to 50 m and 10 −1 to 10 m/s, respectively. A simple model is proposed to evaluate the extent of fractionation as magma traverses cold lunar lithosphere. If Apollo green glasses are primitive and have not undergone significant fractionation en route to the surface, then mean ascent rates of 10 m/s and cracks of widths > 40 m are indicated. Lunar tephra and vesiculated basalts suggest that a volatile component plays a role in eruption dynamics. The predominant vapor species appear to be CO, CO 2, and COS. Near the lunar surface, the vapor fraction expands enormously and vapor internal energy is converted to mixture kinetic energy with the concomitant high-speed ejection of vapor and pyroclasts to form lunar fire fountain deposits such as the Apollo 17 orange and black glasses and Apollo 15 green glass.

Mid-Pleistocene lavas from the Seguam volcanic center, central Aleutian arc: closed-system fractional crystallization of a basalt to rhyodacite eruptive suite

In contrast to adjacent volcanic centers of the modern central Aleutian arc, Seguam Island developed on strongly extended arc crust. K-Ar dates indicate that mid-Pleistocene, late-Pleistocene, and Holocene eruptive phases constitute Seguam. This study focuses on the petrology of the mid-Pleistocene, 1.07–07 Ma, Turf Point Formation (TPF) which is dominated by an unusual suite of porphyritic basalt and basaltic andesite lavas with subordinate phenocryst-poor andesite to rhyodacite lavas. Increasing whole-rock FeO*/MgO from basalt to dacite, the anhydrous Plag+Ol+Cpx±Opx±Mt phenocryst assemblage, groundmass pigeonite, and the reaction Ol+Liq=Opx preserved in the mafic lavas indicate a tholeiitic affinity. Thermometry and comparison to published phase equilibria suggests that most TPF basalts crystallized Plag+Ol+Cpx±Mt at ≥1160°C between about 3–5 kb (±1–2% H2O), andesites crystallized Plag+Cpx+Opx±Mt at ≥1000°C between 3–4 kb with 3–5% H2O, and dacites crystallized Plag +Cpx±Opx±Mt at 1000°C between 1–2 kb with 2–3% H2O. All lavas crystallized at f o 2 close to the NNO buffer. Mineral compositions and textures indicate equilibrium crystallization of the evolved lavas; petrographic evidence of open-system mixing or assimilation is rare. MgO, CaO, Al2O3, Cr, Ni, and Sr abundances decrease and K2O, Na2O, Rb, Ba, Zr, and Pb increase with increasing SiO2 (50–71%). LREE enrichment [(Ce/Yb)n=1.7±0.2] characterizes most TPF lavas; total REE contents increase and Eu anomalies become more negative with increasing SiO2. Relative to other Aleutian volcanic centers, TPF basalts and basaltic andesites have lower K2O, Na2O, TiO2, Rb, Ba, Sr, Zr, Y, and LREE abundances. 87Sr/86Sr ratios (0.70361–0.70375) and ratios of 206Pb/204Pb (18.88–18.97), 207Pb/204Pb (15.58–15.62), 208Pb/204Pb (38.46–38.55) are the highest measured for any suite of lavas in the oceanic portion of the Aleutian arc. Conversely, eNd values (+5.8 to+6.7) are among the lowest from the Aleutians. Sr, Nd, and Pb ratios are virtually constant from basalt through rhyodacite, whereas detectable isotopic heterogenity is observed at most other Aleutian volcanic centers. Major and trace element, REE, and Sr, Nd, and Pb isotopic compositions are consistent with the basaltic andesitic, andesitic, dacitic, and rhyodacitic liquids evolving from TPF basaltic magma via closed-system fractional crystallization alone. Fractionation models suggest that removal of ∼80 wt% cumulate (61% Plag, 17% Cpx, 12% Opx, 7% Ol, and 3% Mt) can produce 20 wt% rhyodacitic residual liquid per unit mass of parental basaltic liquid. Petrologic and physical constraints favor segregation of small batches of basalt from a larger mid-crustal reservoir trapped below a low-density upper crustal lid. In these small magma batches, the degree of cooling, crystallization, and fractionation are functions of the initial mass of basaltic magma segregated, the thermal state of the upper crust, and the magnitude of extension. Tholeiitic magmas erupted at Seguam evolved by substantially different mechanisms than did calc-alkaline lavas erupted at the adjacent volcanic centers of Kanaga and Adak on unextended arc crust. These variable differentiation mechanisms and liquid lines of descent reflect contrasting thermal and mechanical conditions imposed by the different tectonic environments in which these centers grew. At Seguam, intra-arc extension promoted eruption of voluminous basalt and its differentiates, unmodified by interaction with lower crustal or upper mantle wallrocks.

Crustal Processes of the Mid-Ocean Ridge

Independent geological and geophysical investigations of the Mid-Ocean Ridge system have begun to focus on the nature of the magma chamber system underlying its central axis. Thermal models predict the existence of a steady-state chamber beneath a thin crustal lid ranging in thickness from 2 to 13 kilometers. The only aspect of the system that these models fail to account for is the extremely slow spreading rates. Seismological studies reveal the existence of a low-velocity zone beneath segments of the East Pacific Rise, which is thought to correspond to a chamber system having a half-width of approximately 5 to 10 kilometers. These estimates compare favorably with those derived separately through petrological investigations of deep-sea drilling results, various sampling programs, and field and laboratory studies of ophiolites. The chamber is thought to be wing-shaped and to remain continuously open; it is thought to be fed from the center while simultaneously solidifying at the sides as spreading carries the two halves apart. Progressive fractionation occurs by crystal settling coupled with repeated replenishment and magma mixing in an open steady-state system. Near-bottom studies reveal that the zone of extrusion above the chamber is narrow, but its eruptive history is cyclic in nature, in conflict with the predictions of a steady-state model. On-bottom gravity data at 21 degrees N on the East Pacific Rise reveal a negative gravity anomaly that may be related to the uppermost part of the chamber. The anomaly is only 2 kilometers wide and 1 kilometer below the sea floor. This feature may be associated with a short-term upper magma reservoir. The cyclic volcanic activity is directly related to the active phase of hydrothermal circulation responsible for the observed negative thermal anomaly. The volume of water associated with this circulation is equal to the entire ocean volume passing through the accretion zone approximately every 8 million years. This is about 0.5 percent of the world's rivers, but the effective transport rates of elements are comparable to those of rivers in that anomalies for individual elements are frequently between 100 and 1000 times the average river composition. The degree of subsurface dilution determines the final exit temperature and composition of the hydrothermal fluids, ranging from manganese domination at extreme dilution to iron at intermediate levels to sulfide deposition when low dilution occurs. The discovery of massive sulfide deposits on the East Pacific Rise is destined to have a profound impact on our understanding of ore-forming processes. Whether it will have any economic significance remains to be seen.