Compositional variation of density and seismic velocities in natural peridotites at STP conditions: Implications for seismic imaging of compositional heterogeneities in the upper mantle

Abstract

[1] Densities and elastic properties of solid natural spinel-and garnet-peridotite samples (n = 133) at standard temperature (T ) and pressure (P) (STP) conditions were calculated for compositions ranging from Mg# (100 Â Mg/(Mg + Fe)) of 86-94. The physical properties were used to investigate how natural compositional variations control density and seismic velocity. A corresponding set of compositional derivatives (d/dMg#) of density and seismic velocity is provided. Because the P and T derivatives of elastic moduli are very similar for different compositional end-members of peridotitic minerals, the variation of elastic moduli with Mg# at STP conditions holds at elevated P and T. Increased Mg# leads to a significant increase in V S because of the sensitivity of mineral shear moduli to this parameter (dV S /dMg# = 0.0143 ± 0.0009 km s À1). In contrast, the compressional wave velocity (V P ) is insensitive to Mg# and, instead, correlates weakly with increasing olivine abundance at STP conditions. The ratio V P /V S therefore exhibits a significant negative correlation with Mg# (d(V P /V S )/dMg# = À0.00407 ± 0.00038). Because the temperature dependency of V P /V S is small (<$0.04%/100°C) compared to the compositional dependency (1.7%/Mg# unit), the variation in V P /V S is a fairly robust measure of compositional variation even when temperature varies. Finally, a new density versus Mg# parameterization is derived. Combined with a compilation of bulk-rock Mg#s of peridotite xenoliths from cratonic lithospheric mantle, it is shown that the intrinsic density of cratonic mantle balances to within error its negative thermal buoyancy imposed by its cooler thermal state relative to upwelling asthenospheric mantle. INDEX TERMS:1025 Geochemistry: Composition of the mantle; 3909 Mineral Physics: Elasticity and anelasticity; 7207 Seismology: Core and mantle; 8120 Tectonophysics: Dynamics of lithosphere and mantle-general; KEYWORDS: xenolith, peridotite, seismic velocity, density, lithosphere, tectosphere Citation: Lee, C.-T. A., Compositional variation of density and seismic velocities in natural peridotites at STP conditions: Implications for seismic imaging of compositional heterogeneities in the upper mantle,

CrustalVSstructure in northwestern Canada: Imaging the Cordillera-craton transition with ambient noise tomography

“…[45] In the upper mantle, temperature is the primary cause of perturbations in wave speed at a fixed depth; variability in rock and mineral composition has a smaller effect [e.g., Lee, 2003;Shapiro and Ritzwoller, 2004;Faul and Jackson, 2005]. Recent seismic images of the upper mantle across northwestern Canada have been obtained from teleseismic body waves [e.g., Mercier et al, 2008Mercier et al, , 2009] and surface waves [Frederiksen et al, 2001; van der Lee and Frederiksen, 2005;Nettles and Dziewonski, 2008].…”

“…These high velocities are primarily due to cold temperatures, although the depleted composition of cratons may also contribute [e.g., Jordan, 1978;Lee, 2003]. Young orogens, on the other hand, are typically low-velocity features in the upper mantle.…”

Seismic structure of Iceland from Rayleigh wave inversions and geodynamic implications

“…This depleted, less dense layer beneath Iceland hotspot was proposed by Jordan [32] in 1979. Because melt generated in the plume conduit migrates upward, the residual mantle left behind becomes more depleted and Mg-rich, and will be associated with a higher velocity [32][33][34]. The fast anomaly we imaged is centered at ~135 km depth, while the initial depth of melting beneath Iceland is estimated at ~110 km depth from geochemistry data [35], which is adopted in geodynamical modeling [29][30][31].…”

Imaging teleseismic P to S scattered waves using the Kirchhoff integral

“…In the presence of modest amplitude random noise the small amplitude event is still well imaged. and Mooney [1995] to the top of the transition zone to account for the reduced Mg# of depleted cratonic lithosphere rocks [Lee, 2003]. Right: The cratonic lithosphere has a significantly lower V p /V s ratio than the IASP91 model.…”

“…Crustal velocities were taken from the crustal model of Christensen and Mooney [1995] for cratons ( Figure 6). A second modification was to reduce the compressional velocities in the mantle between the crust and the 410-km discontinuity to account for a depleted cratonic root ( Figure 6) [Lee, 2003;Niu et al, 2004]. Iron depletion of the upper mantle during craton formation not only increases the Mg# of the rocks forming the upper mantle, but also affects the V p /V s ratio largely by reducing compressional velocities relative to S for a fertile mantle.…”