HR: 09:30h
AN: S41E-07    [Abstracts]
TI: Inference of upper-mantle density structure from seismic velocities
AU: * Nettles, M
EM: nettles@eps.harvard.edu
AF: Dept. of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138
AU: Dziewonski, A M
EM: dziewons@eps.harvard.edu
AF: Dept. of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138
AB: The inverse problem for the determination of density structure from perturbations in the gravity field is highly nonunique. The combination of gravity data and other observables can, however, be used to make inferences about the Earth's density structure in three dimensions. We use the three-dimensional shear-wave velocity model of Nettles and Dziewonski (2005) to make a forward prediction of the Earth's gravity field using simple assumptions about the relationship between perturbations in shear velocity and density. A scaling factor f=0.25 relating perturbations in shear velocity and density (δ/ρρ = f · δv_S/v_S) is determined empirically by comparison of observed variations in shear velocity in oceanic regions with density variations predicted from a simple model of conductive cooling. This value agrees well with f=0.27 based on the laboratory results of Jackson et al. (1992). The observed gravity signal in the oceans is explained well by this simple thermal-scaling approach. Behavior in some continental regions, such as the Basin and Range and the East African rift zone, is found to be similar to that in the oceans: the high topography in these regions appears to be supported by hot, low-density mantle underneath, a result also found by Kaban and Mooney (2001) for the Basin and Range. A velocity-to-density scaling relationship based only on thermal considerations is clearly inadequate in regions of continental craton, where such scaling leads to unrealistically large perturbations in the predicted gravity field. This result suggests that non-thermal effects must counteract the high density that would occur due to thermal effects alone, consistent with the suggestion of Jordan (1975) and other workers that density increases due to cool temperatures in the continental roots must be balanced by density decreases due to compositional variations. Using the compositional derivatives for density and shear velocity with respect to Mg# determined by Lee (2003), and an assumption of neutral buoyancy in the continental roots, yields perturbations in temperature in agreement with those obtained by Artemieva and Mooney (2001) based on heat-flow and heat-production data and perturbations in Mg# that are generally consistent with values observed for cratonic mantle xenoliths.
DE: 7208 Mantle (1212, 1213, 8124)
DE: 7270 Tomography (6982, 8180)
DE: 8103 Continental cratons
DE: 8122 Dynamics: gravity and tectonics
SC: Seismology [S]
MN: Fall Meeting 2005