Barbara Romanowicz, "Can we resolve 3D density heterogeneity in the lower mantle?". Geophys. Res. Lett., 28, 1107-1110 (2001).
Download pdf reprint
We reexamine the possibility of resolving 3D density structure in the mantle,
by performing test inversions using a dataset of well constrainted degree 2 splitting coefficients. We show that degree 2 in Vs and Vp can be resolved, and confirm that dlnVs/dlnVp >2.5 at depths > 2000 km, indicating a chemical component to large scale heterogeneity. Models with rho positively correlated with Vs in the deepest mantle fit the data at least as well as those in which density and velocities are anti-correlated. Howver, some rough bounds on dln rho/dln Vs can be inferred. Models obtained from geodynamic considerations are acceptable, in contrast to models with large negative(<-0.3) or positive (>0.6) dln rho/dlnVs throughout the lower mantle. The most robust feature is a marked
increase in dln rho/dln Vs at the top of the lower mantle reaching a maximum
around 1500 km.
Download pdf reprint
The resolution of the 3-D density structure in the Earth s deep interior has long eluded geoscientists. High-quality data from digital seismic instruments emplaced this decade have renewed interest in the measurement of low-frequency Earth normal modes with the goal of extracting heterogeneous density structure. Here we perform a series of synthetic experiments aimed at investigating the resolution of lateral variations in the mantle from normal-mode spectral data. Contamination effects between seismic velocities and density are examined in two ways: (1) by using resolution matrices computed from data kernels and (2) by inverting synthetic spectra computed from realistic input Earth models. The first type of experiment assumes that the non-linearity of the inverse problem is weak. No such assumptions are necessary in the second type of tests which, nevertheless concur in their results with the former. These synthetic tests indicate that density structure retrieved from presently available normal-mode data is not reliable. Contamination of seismic velocity structure into the density model space produces patterns that resemble those obtained from inversions of real data. The resulting density models appear to be dependent on a combination of the starting velocity models and the model parametrization.
For all enquiries, please contact Barbara Romanowicz:
barbara@seismo.berkeley.edu
or Chaincy Kuo:
chaincy@cal.berkeley.edu