The travel times and amplitudes of various seismic phases yield detailed information on the deepest portions of the Earth's mantle may be extracted. This information relates directly to important geophysical phenomena, such as modes of mantle circulation and extraction of heat from the molten liquid core to the planet's surface. Below I discuss a research program focusing of different structural characteristics of the lower mantle and D" (the lowermost few hundred km of the mantle) and the seismic waves that were used to study D" heterogeneity, anisotropy, and ultralow velocity zones.
Most recently, I have been finishing the compilation of a data set of S-SKS and SKKS-SKS differential times for use in an inversion for lowermost mantle heterogeneity (with Ban Kuo, Academia Sinica, Taiwan). Our solution models looks much like that obtained from whole mantle global tomographic inversions.
At smaller scales (e.g., 500 km), coherent patterns of heterogeneity are observed from ScS-S times southeast of Hawaii (ongoing work with Sara Russell and Thorne Lay, UC Santa Cruz). Such forward modeling yields remarkable results: at smaller and smaller scales, systematic/coherent patterns of heterogeneity emerge (as also seen by UCB colleagues Breger and Romanowicz). Our exciting task is now to interpret such patterns in terms of lower mantle dynamics, temperature, and chemistry.
Understanding the nature of anisotropy in the D" region may answer important questions regarding the dynamics of lower mantle material flow and strain fields. Unfortunately, many uncertainties exist. Nonetheless, we have been successful at mapping out patterns in anisotropy as well as conducting synthetics tests on various model features.
With Russell and Lay, we've detected a systematic change in the fast polarization direction for split ScS waves over short lateral scales in D", just SE of Hawaii, and proposed the source of this anomaly as being related to the root of the Hawaiian plume. Currently we are combining older analog and modern broadband digital data to analyze ScS behavior beneath the Caribbean. Our work (along with Raymond Jeanloz, UCB, Quentin Williams, UC Santa Cruz, and Jeroen Ritsema, Caltech) has focused on differences between the patterns of anisotropy beneath proposed regions of upwelling and downwelling currents in the overlying mantle.
Work with Don Helmberger (Caltech) has resulted in the detection of very thin zones (of order 10 km) of ultralow velocities at the base of the mantle. P wave speeds are reduced in this layer by 10 30 zone (ULVZ) as being due to partial melt of lowermost mantle material. Patches of ULVZ material are strongly correlated spatially with hotspots at the Earth's surface, a result we use to argue for a CMB source to hotspots through whole mantle traversing plumes (with Williams and Justin Revenaugh, UC Santa Cruz).
The past several years has produced a rich variety of information regarding the deep Earth and its processes, allowing us to summarize our views of how the deep Earth 'works' (Lay et al., 1998). While this process is very fun and exciting, we have also realized the importance for exploring more deeply some of the uncertainties involved in the seismic modeling. With D. Helmberger, we've shown that strong trade-offs exist in ULVZ modeling. R. Jeanloz and I are exploring the possibility that the ULVZ could actually be an outermost core phenomenon, with a 1.5-2 km slush layer existing under CMB "bumps", as 'inverse lakes'. Also, T. Lay and I have recently shown that D" anisotropy can affect seismic waves that enter the core, such as SKS and SKKS. The following year will be a continuation of these investigations and collaborations.
Lay, R., Q. Williams, and E. J. Garnero, The core-mantle boundary layer and deep Earth dynamics, Nature, 392, 461-468, 1998.