Thermal structure of continental upper mantle inferred from S-wave velocity and surface heat flow (original) (raw)
Related papers
Shear wave velocity, seismic attenuation, and thermal structure of the continental upper mantle
Geophysical Journal International, 2004
Seismic velocity and attenuation anomalies in the mantle are commonly interpreted in terms of temperature variations on the basis of laboratory studies of elastic and anelastic properties of rocks. In order to evaluate the relative contributions of thermal and non-thermal effects on anomalies of attenuation of seismic shear waves, Q −1 s , and seismic velocity, V s , we compare global maps of the thermal structure of the continental upper mantle with global Q −1 s and V s maps as determined from Rayleigh waves at periods between 40 and 150 s. We limit the comparison to three continental mantle depths (50, 100 and 150 km), where model resolution is relatively high.
Correlation between the shear-speed structure andthickness of the mantle transition zone
Physics of the Earth and Planetary Interiors, 2003
The 410 and 660 km seismic discontinuities that bound the mantle transition zone (TZ) are attributed to phase transformations in olivine structure. This implies that variations in TZ thickness (H TZ ) should correlate with those in TZ temperature. Pertinent seismic evidence has so far been ambiguous, however. We measure converted-wave (P ds) differential times t diff = t P 660s − t P 410s in SE Asia and Australia and compare them with S-velocity (β TZ ) estimates from regional tomographic models. Both t diff and β TZ vary on a scale of a few hundred kilometers. Inferred variations in H TZ are up to ±30 km over length scales larger than 500 km, implying ±200 K thermal heterogeneity if the effect of composition can be neglected. t diff and β TZ correlate strongly; the linear dependence of H TZ on the average temperature within the TZ is consistent with olivine Clapeyron slopes. We also show that this relationship holds on a global-scale as well, provided that the scalelengths and uncertainties of the variations in t diff and β TZ are taken into account. These results confirm that the transformations in olivine structure give rise to the 410 and 660 km discontinuities globally.
Journal of Geophysical Research, 1998
A more general expression for the mantle vorticity equation is proposed for convection using axisymmetrical spherical geometry. Both the main mantle phase changes and radial and lateral variations of viscosity due to temperature and pressure. Four series of computations have been performed with (1) both the latent heat releases of the 400 km exothermic and the 670 km endothermic ph .ase change and uniform and constant mantle viscosity; (2) the 670 km phase change alone and viscosity jumps of 10 or 30 between upper and lower mantle phases; (3) the 670 km endothermic phase change, a viscosity contrast of 30, and temperature and pressure dependent viscosity law; and (4) both 400 km and 670 km phase changes, a viscosity jump of 30, and a temperature and pressure dependent viscosity. The 400 km exothermic phase change modifies the global structure from partly layered to whole mantle convection. This effect is opposite to the effect obtained by increasing the viscosity jump at 670 km. However, both effects induce unrealistic thermal behavior which will not appear with temperature dependent laws for viscosity. The mantle avalanches which suddenly inject huge quantities of cold material into the lower mantle have effects at the surface and at the coremantle boundary (CMB). They induced heat flow crises which explain the huge volcanic events, high rates of mid-oceanic ridge accretion, and periods of low-frequency magnetic reversal. The surface heat flow proceeds directly from the upper mantle return flow along with the avalanches. The temperature dependent viscosity tends to decrease the strength of the avalanches. The bottom heat flow and the birth of CMB plumes may be considered as the consequences of cold upper mantle material arrival at the CMB. The lower mantle and the upper mantle transit times depend on the thickness of upper and lower mantles but also on the phase changes and on the viscosity. The CMB and surface perturbations may be simultaneous (to a few tens of million years).The temporal evolution of the convection pattern during an avalanche allows us to propose self-consistent mechanisms for slab migration above the 670 km discontinuity for the birth and disappearance of ridges, the rising of powerful plumes from the CMB, and the creation of low-viscosity zones which may act as a lubricant under continents for fast migration. These results show that the main mantle phase changes, combined with temperature and pressure dependent viscosity, induce convective behavior which provides an explanation for most of the past and present large-scale dynamic behavior of the Earth's global tectonics. equations (including the thermal equation). In the first approach, the density anomalies of kinematic models are computed from lateral seismic velocity anomalies which may be considered as a rough map of the lateral thermal variations within the mantle. From the first generation of these dynamic models [see, e.g., Richards and Hager, 1984; Ricard et al., 1984] to the most recent models, several improvements have been taken into account such as the compressibility of mantle [Thoraval et al., 1994], the motion of surface plates [e.g., Ricard and Vigny, 1989], and the effects of the main mantle discontinuities [Thoraval et al., 1995]. In spite of a still present debate about the amplitude of the dynamic surface topography, these kinematic models mainly agree that an abrupt mantle viscosity jump of about 30 between the upper and the lower mantles is required to retrieve the main pattern of the geoid. The second approach consists of solving the full set of convection equations. This allows us to gain information about the temporal behavior of the mantle and therefore about the thermal evolution of the Earth. Recently, these models have proved the chaotic behavior of mantle convection which prevents geophysicists from retrieving the exact pattern of temperature within the mantle [Machetel and Yuen, 1986, 1989]. However, this approach is able to check the geophysical relevance of the main model assumptions. 4929 4930 BRUNET AND MACHETEL: MANTLE CONVECFION AND LARGE-SCALE 2F_,C-TONIC Christensen and Yuen [1984, 1985] were the first to compute the embedding behavior of the endothermic phase change. New interest has been stimulated with the work by Machetel and Weber [1991], which showed that an intermittent layering, punctuated by sudden avalanches of upper mantle material sinking deeply and rapidly into the lower mantle, may be the prevailing regime within the mantle. This behavior, which allows us to reconcile most of the geophysical arguments in favor of either a whole mantle or a layered convection, is closely related to the negative Clapeyron slope of the 670 km depth phase change. Since then a number of studies have been devoted to this question. The effects of the 670 km endothermic phase change on the structure of mantle convection have been widely documented by numerical models. They have been run in twoand three-dimensional (2-D and 3-D) Cartesian geometry [i.e., Steinbach and Yuen, 1992; Honda et al., 1993; Steinbach and Yuen, 1994; Yuen et al., 1995] and in spherical geometry with the axisymmetrical assumption [Machetel and Weber, 1991; Peltier and Solheim, 1992; Bercovici et al., 1993; Machetel, 1993; Solheim and Peltier, 1994a,b] and with full 3-D 15917, 1994.
Generation of plate-tectonic behavior and a new viscosity profile of the Earth's mantle
2003
This paper reports a series of compressible spherical-shell convection calculations with a new viscosity profile, called eta3, that is derived from PREM and mineral physics. The viscosity profile displays not only a high-viscosity lithosphere and a viscosity hill in the central region of the lower mantle of the Earth but also a prominent high-viscosity transition layer inferred to arise from a high garnet content. Moreover, there is not only the usual asthenosphere but also a second low-viscosity zone just below the 660-km discontinuity. We introduced a viscoplastic yield stress and obtained plate-like movements near the surface. A variation of the parameters revealed a Rayleigh-number-yield-stress area where the plate-tectonic character of the solution is stable and pronounced. Runs with eta3 but without any yield stress show networks of reticularly connected very thin sheet-like downwellings but, of course, no plates. For calculations with eta3 plus yield stress, the distributions of the downwellings are more Earth-like.
Annu. Rev. Earth Planet. Sci., 2008
Rock-mechanics experiments, geodetic observations of postloading strain transients, and micro-and macrostructural studies of exhumed ductile shear zones provide complementary views of the style and rheology of deformation deep in Earth's crust and upper mantle. Overall, results obtained in small-scale laboratory experiments provide robust constraints on deformation mechanisms and viscosities at the natural laboratory conditions. Geodetic inferences of the viscous strength of the upper mantle are consistent with flow of mantle rocks at temperatures and water contents determined from surface heat-flow, seismic, and mantle xenolith studies. Laboratory results show that deformation mechanisms and rheology strongly vary as a function of stress, grain size, and fluids. Field studies reveal a strong tendency for deformation in the lower crust and uppermost mantle in and adjacent to fault zones to localize into systems of discrete shear zones with strongly reduced grain size and strength. Deformation mechanisms and rheology may vary over short spatial (shear zone) and temporal (earthquake cycle) scales.
Journal of Geophysical Research, 1997
Joint inversions of seismic and geodynamic data are carried out in which we simultaneously constrain global-scale seismic heterogeneity in the mantle as well as the amplitude of vertical mantle flow across the 670 km seismic discontinuity. These inversions reveal the existence of a family of three-dimensional (3-D) mantle models that satisfy the data while at the same time yielding predictions of layered mantle flow. The new 3-D mantle models we obtain delnonstrate that the buoyancy forces due to the undulations of the 670 km phase-change boundary strongly inhibit the vertical flow between the upper and lower mantle. The strong stabilizing effect of the 670 km topography also has an important impact on the predicted dynamic topography of the Earth's solid surface and on the surface gravity anomalies. The new 3-D models that predict strongly or partially layered mantle flow provide essentially identical fits to the global seismic data as previous models that have, until now, predicted only whole-mantle flow. The convective vertical transport of heat across the mantle predicted on the basis of the new 3-D models shows that the heat flow is a rninimum at 1000 km depth. This suggests the presence at this depth of a globally defined horizon across which the pattern of lateral heterogeneity changes rapidly. ing used to draw a variety of inferences about the properties of the thermal convection process in the mantle [e.g., Hagcr and Clayton, 1989; Forte and Pcltier, 1991; King and Masters, 1992; Forte et al., 1993a; Kogan and McNutt, 1993; Phipps Morgan and Shearer, 1993; Jordan et al., 1993; Woodward et al., 1994]. One of the main outstanding issues of geodynamics that has yet to be resolved by these tomography-based studies is whether the mantle-convective flow is layered or not. This issue centers on whether the global seismic discontinuity near 670 km depth [e.g., Dziewonski and Anderson, 1981; Shearer, 1991], generally assumed to be a manifestation of the pressure-induced transformation of olivine spinel to its postspinel phases [e.g., Anderson, 1967; Jeanloz and Thompson, 1983; Ito and Takahashi, 1989], acts as a barrier to the vertical convec-tive transport ' of mass and heat between the upper and lower mantle. If the convective circulation in the upper and lower mantle were segregated, then the bulk chemical composition of these two regions may be distinctly different. It has thus been suggested that the seismic contrast near 670 km depth may be in part due to an abrupt change in chemical composition [e.g., Liu, 1979; Anderson, 1981; Jeanloz, 1991]. Some of the main arguments proposed either in favour of, or against, layered thermal convection in the mantle have been reviewed by Jordan et al. [1989], Jeanloz [1989], and Peltier et al. [1989]. Interest in the possibility of layered mantle convection has received much impetus from a recent series of numerical simulations of thermal convection that have shown varying degrees of inhibition of vertical flow across the spinel-postspinel phase change boundary. A detailed review of these phase-change-modified thermal convection simulations in 3-D geometry has recently been presented by Tackleit [1997]. A comprehensive discussion of the important mineralogical and theological consequences of mantle phase transitions is given by Ka•ato [1997]. In view of the implications of layered thermal convection for the thermal and chemical evolution of the Earth [e.g., Jeanloz and Richier, 1979; Peliier el al., 1989; $olheim and Peltier, 1993; Stein and Holmann, 17,981 ], there have been several efforts to verify the plausibility of the layered convection simulations using recent seismic tomographic inferences of 3-D mantle structure [e.g., Forte et at.Thoraval et al. , 1995]. The importance of these independent tests of layered mantle flow stems, in part, from the concern that the many simplifications employed in the numerical convection simulations may limit their relevance to the actual state of thermal convection occurring in the Earth's mantle.
What Can Seismology Say About Hot Spots?
2015
Seismology offers the highest-resolution view of mantle structure. In the decades since Morgan [1971] first proposed deep-mantle plumes, seismologists have used increasingly sophisticated methods to look for evidence of such structures, but so far they have had little success. This abstract outlines the relevant seismological methods for non-specialists and summarizes the current state of knowledge about structure beneath hot spots, to set the stage for the seismological component of this conference. Factors Affecting Seismic-Wave Speeds Direct thermal effect – If thermal plumes exist in the mantle, they would have lower seismic wave speeds than their surroundings. In the upper mantle, a 100 K temperature rise lowers the compressional-wave speed, VP, by about 1%, and the shear-wave speed, VS, by about 1.7%. In the deep mantle, this effect is several times weaker. The temperature anomalies proposed for plumes are about 200 to 600 K. Indirect thermal effect – Temperature variations al...