Current Research Projects:

Plate-Mantle Interaction and Forces that Drive the Plates >>Papers. Although the concept of plate tectonics is universally accepted, it is incomplete in that there is no agreement on the mechanism that drives the plates. It has been suggested that they are driven from the side, that is pulled by subducting slabs and pushed by negative buoyancy forces from subsidence of oceanic plates ("ridge-push" force). Another view is that plate motion is driven by mantle convection. The motion of the plates can be explained by either of these concepts. Seismic anisotropy, on the other hand, records deformation at depth, and may in principle determine the sense of relative motion between plates and deeper mantle. We suggest that the pattern of deformation in the deep continental roots under the shield areas give important insights into the question of driving forces, since they represent mechanical obstacles in the relative motion between plates and deeper mantle. Seismic anisotropy may allow to determine the sense and also the level of shear stress at the base of the plates. The technique is based on the analysis of teleseismic P-wave delays. One of the results of this study is that the deeper mantle helps to drive the North American plate. These and related questions were discussed in a union session at the fall 2000 AGU meeting. Recently Peter Bird confirmed the view that the mantle under North America helps to drive plate motion.

Mechanics of a Plate Boundary System (California) >>Papers. Some important questions in the context of plate boundary deformation are: How is plate boundary deformation accomodated at depth? What is the role of the lower crust in crustal deformation (detachment, block rotation, depth extension of faults)? Does it decouple the upper crust from the mantle? Do crust and mantle share a coherent deformation? How are driving forces for crustal deformation applied, e.g., do they act from the side or from below? Is the lower crust mechanically weak or strong? And how does the weakness/strength of the lower crust relate to the weakness of major faults in the upper crust? Is their long-term strength as low as the heat flow paradox suggests? And how are they loaded before there is a large earthquake? What role do crustal fluids play in this? We have developed a new tool for studying stress at depth, which allows us to address these questions? It uses a proxy for the shear traction on horizontal planes (e.g., the free surface has very small shear traction). In Northern California, the surface and also the base of the seismogenic crust show up with low shear traction. This suggests that the lower crust is mechanically weak, or that perhaps it relaxes with time. Interestingly, this behaviour occurs on both locked and creeping sections in Northern California. This technique also allows important insight into depth-dependent properties of fault zones.

Large-Scale Earthquake Relocation (California) >>Papers. Newly developed high-resolution images of seismicity on faults within the San Andreas system reveal structure that was previously obscured by earthquake location errors. The techniques combine ordinary travel time picks with high-precision phase correlation measurements. In particular, hypocenters on the Hayward, Calaveras and San Andreas faults show that seismicity is highly organized in both space and time. Common features from all of these faults include: 1) a narrow (25-100 m) width of the fault core; 2) approximately horizontal streaks of hypocenters along the fault plane; 3) multiple recurrence of earthquakes of the same size at precisely the same spot on the fault (multiplets); and 4) regions devoid of earthquakes (holes). The streaks have typical lengths of up to 5 km. The holes are perhaps the most intriguing feature, as in the long term, the fault must slip on these zones, either in earthquakes or by steady or episodic creep. If such holes represent likely nucleation regions of future earthquakes they suggest opportunities for focused natural laboratory experiments in the context of the Plate Boundary Observatory. This is/was a collaborative effort of groups at Stanford University and the USGS at Menlo Park.  For a text describing the potential of these techniques in the context of global earthquake relocation see the IRIS-Newsletter.

Time Dependent Earth Properties >>Papers. The shape of Earth's surface varies over geological time scales and similarly does the structure of Earth's interior. But are there measurable changes also on time scales of the duration of a scientific project, say within 1-3 years? Our primary measurements are seismic waves, and for waves that propagate along the same paths at different times one can measure temporal changes in the medium. We have striking observational evidence that the medium seen by seismic waves is altered during large earthquakes, e.g. after the 1989 Loma Prieta earthquake in Central California (medium change over the 6 years following the event) and it slowly recovers in the following years. Such temporal changes are primary indicators for the dynamics of fault zone behaviour, e.g. fluid migration in the vicinity of the earthquake rupture, stress propagation, and perhaps fault corrugation. If we can measure temporal changes very accurately, we may use the technique to measure the buildup of mechanical stress at seismogenic depths, that is where the earthquakes occur, and to measure fault zone corrugation in situ. Such techniques can also be used for monitoring oil and gas reservoirs. Beside natural sources we also use man-made sources, which run continuously and emit waves that are measured to distances of up to several hundred km.

Plumes and Mantle Dynamics >>Papers. There are a number of unresolved yet fundamental questions about the nature of mantle upwellings or "plumes" under hotspots. Specifically, can one distinguish the notion of plumes, i.e., material upwelling from parts of the deeper mantle, from alternative models as a source for mantle hotspots? What causes topographic swells above hotspots? And what is the geometry of upper-mantle flow in a region where a plume impinges upon the base of the lithosphere? To help answer these questions, it is important to directly resolve aspects of upper-mantle flow around mantle upwellings. This is possible by making measurements of shear-wave splitting for seismic waves that travel through the upper mantle, since flow-related deformation generally causes seismic anisotropy and the geometry of the anisotropy is reflected in these measurements. In particular, shear-wave splitting for waves that travel through the upper mantle is a useful tool for this purpose, since the polarization direction of the faster shear-wave documents the near-horizontal flow. A guiding hypothesis for how mantle upwellings interact with the base of the lithosphere is the "parabolic flow model". A mantle plume rising beneath a moving plate induces a pattern of upper-mantle anisotropy in which the fast directions align in an approximately parabolic pattern in the horizontal plane centered above the plume. The areas which we are working on include Hawaii, the Snake River Plain (Yellowstone), and the Eifel area. We also address plume-related tectonic structure in Western US.