The Ninth Annual IRIS Workshop was held June 8-12th at the Beaver Run Resort in Breckenridge Colorado. More than 250 people attended the Workshop and participated in various activities, including four fieldtrips of hiking - Geology of the Northern Rio Grande Rift lead by Colorado School of Mines; a hike to the snow line lead by T. Wallace; white-water rafting lead by the Breckenridge Outdoor Education Center, who also lead a canoeing trip on the ecology of Lake Dillon. Even rain did not dampen the spirts of the participants.
IRIS introduced our new Education and Outreach Program by providing a one day short course on "Seismologists Learning to Teach the Teachers" (described in previous article). There was also an evening of "the Great Debate" where four members of the community debated the relative importance of science education from K-12 to the graduate level, and public education.
 | The Great Debate, lead by Gregory van der Vink. The participants - Adam Dziewonski (K-12), Steven Bohlen (undergraduate), Ian MacGregor (graduate), and Kaye Shedlock (public education) - debated the relative importance of science education at various grade levels. |
Tom Jordan led off the session with a talk entitled "New Constraints on the Structure and Evolution of the Continental Tectosphere" (co-authored by Jim Gaherty). Key points included xenolith evidence for chemical depletion in the tectosphere/mantle root (resulting in positive chemical buoyancy that competes with negative thermal buoyancy) and for roughly similar ages between archean cratonic crust and the underlying mantle. Jordan integrated these results into a model in which the tectosphere formed by multiple episodes of advective thickening prior to its stabilization. In a refinement of the original tectosphere hypothesis, Jordan used waveform inversion results from Australia to argue that the Lehman discontinuity occurs at 190 to >250 km depth, marking a transition from more rigid anisotropic tectosphere at shallower depth to more mobile, isotropic tectosphere at greater depth. In "The Thermal and Dynamical Evolution of Sub-continental Lithosphere," Chris Kincaid (with Paul Silver as co-author) explored a wide variety of geodynamical calculations that emphasized the large impact of continental root buoyancy and viscosity on mantle flow patterns. The stability of mantle roots over long times scales (> 1 Ga) appears to be controlled by root density and thermal and chemical contributions to root viscosity, and in models where downwelling does occur, it may be localized at root margins as well as beneath root interiors. In numerical modeling of the effects of orogenesis, Kincaid showed that for high values of lithospheric stress (> 200 MPa), viscous heating during orogenesis can substantially increase temperatures in the sub-continental lithosphere, resulting in slower seismic velocities.
Turning the session´s focus to the question of large-scale continental root morphology, Rob van der Hilst presented "Constraints on Deep Continental Structure from Broadband Seismic Imaging." Using data recorded by the SKIPPY project that spanned the Australian continent with a series of broadband arrays, van der Hilst showed velocity images obtained by inverting fundamental and higher mode Rayleigh and body waves. Key features included fast velocities to depths of 300 km or more beneath the Precambrian craton in western and central Australia and thinner lithosphere with a strong low velocity zone in the 140-200 km depth range beneath Phanerozoic eastern Australia, as well as similar strong velocity gradients across root margins in North America (Van der Lee and Nolet) and Europe (Zielhuis and Nolet). Van der Hilst also raised the idea that phase picks from portable arrays be routinely reported to data centers such as the ISC and NEIC, thus offering better coverage in the global travel-time dataset for earthquake location and travel-time inversions.
The next two talks focused on crust and mantle structure across the boundary between craton and orogenic belt in the western U.S. In "Exploring the Depth of Continental Tectonics in the Western United States," Anne Sheehan (with Ken Dueker as co-author) reviewed the numerous and dense PASSCAL experiments that have been conducted across the Wyoming Craton, Rocky Mountains, Colorado Plateau, Basin and Range and adjacent regions over the last several years. One fundamental result from crustal models in the Southern Rocky Mountains, Sierra Nevada and Snake River Plain is that simple crustal roots are insufficient to isostatically support the observed surface topography, arguing for additional support from mantle structure and processes. Sheehan also explored the "410" km and "660" km mantle discontinuities using stacks of Ps conversions recorded by PASSCAL arrays in the Snake River Plain and Rocky Mountains. The "410" and "660" discontinuities have significant topography (15 - 40 km) over short length scales and the "410" and "660" topographies are not simply correlated, arguing that the topography is due to small-scale thermal and/or chemical heterogeneity, rather than thermal anomalies that are vertically coherent through the transition zone. Tim Henstock (with Alan Levander as co-author) then delivered "Contrasting Phanerozoic and Archean Mantle in Western North America: The Deep Probe Experiment." Deep Probe was a 1995 active source deployment that reached 3000 km from southern New Mexico to Great Slave Lake, Canada, crossing from Phanerozoic orogen in the south to Archean craton north of Wyoming. Henstock showed data collected from various shotpoints that reveal fundamental differences in crust and mantle structure between the Colorado Plateau and Wyoming craton. For instance, the mantle beneath the Wyoming craton is high velocity and contains a positive velocity gradient to depths of 150 km or more. However, the mantle beneath the more recently tectonized Colorado Plateau is lower velocity and contains a low velocity zone with a velocity minimum at roughly 80 km depth, a structure reminiscent of young oceanic lithosphere.
In the final talk of the session, "Crustal Reworking During Orogeny: The Nanga Parbat Seismic Experiment," Anne Meltzer focused attention on the effects of orogenesis at shallower depths and more detailed scales. The Nanga Parbat PASSCAL experiment was part of a broad multi-disciplinary effort and involved the deployment of 10 broadband and 50 short period stations in the dramatic topography on and around the Nanga Parbat massif in northeast Pakistan. The age of metamorphism on the massif is very young, and, along with rapid rates of exhumation and young intrusive rocks, suggests very anomalous thermal structure, an elevated geothermal gradient in particular. High rates of local seismicity were recorded during the deployment, and initial event locations show that seismicity is restricted to shallow depths (roughly 8 km or less), providing constraints on the brittle-ductile transition within the crust. Future analyses will include tomographic imaging of the velocity and attenuation structure beneath the massif.
The Tuesday morning session featured a variety of talks concerning the dynamics and upper mantle structure of mid-ocean ridges and continental rifts, lead by Doug Wiens. Don Forsyth presented results from the recent MELT experiment along the East Pacific Rise. This experiment consisted of 51 ocean bottom seismographs (OBS) deployed in two arrays spanning an 800 km region near the spreading center. The results show no evidence of a narrow zone of low velocities near the ridge crest that are predicted by models of active dynamic flow. Melt appears to be broadly distributed near the ridge crest, particularly towards the west, which shows lower seismic velocities and higher topography than the eastern flank. Shear wave splitting and Rayleigh wave phase velocities show pronounced anisotropy, with fast axes parallel to the spreading direction.
Doug Wiens described results from a combined land-sea experiment in the Tonga-Fiji region that used 12 IRIS-PASSCAL land broadband seismic stations and 30 OBSs to image the Lau backarc spreading center and the subducting Tonga slab. Waveform inversion shows exceptionally low seismic velocities at depths of 30-90 km beneath the Lau spreading center, consistent with a zone of magma production. Seismic tomography shows the slow velocities associated with the spreading center are separated from those associated with the Tonga magmatic arc at shallow depths, but merge at depths greater than 100 km, suggesting that slab signatures in backarc magmas may originate through interchanges at greater depths. Some slow velocity signature of the backarc extends to depths of 400 km, possibly due to slab dewatering.
In recent years there have been several different large-scale experiments in Iceland. Guust Nolet described possible ways to study the interaction of a hotspot with a mid-ocean ridge. Seismic tomography images the slow velocity anomaly of the Iceland hotspot extending down into the mantle. The depth extent of the hotspot anomaly is difficult to resolve. One possible way to obtain better resolution of the hotspot anomaly is to model body wave amplitude anomalies caused by focusing. Andy Nyblade continued the discussion by addressing seismic velocity anomalies in the upper mantle beneath the East African Rift. A plume model has been proposed to explain low upper mantle velocities and the volume of volcanics in the rift. However, results from a PASSCAL experiment in Tanzania show the craton is intact to depths of at least 200 km, showing that the mantle has not been disturbed in this region.
Jason Phipps Morgan described his model for the role of the asthenosphere in global mantle flow. He proposed that the asthenosphere is fed by upwelling hot mantle from plumes. The asthenospheric material may be incorporated into lithospheric plates at spreading centers far from the original hotspot source. This model is consistent with the observed dynamic topography and geochemical evidence. Marc Spiegelman completed the discussion by asking whether geochemistry can tell us anything useful about ridge dynamics. Not surprisingly, he concluded that it can, and described how the observed geochemistry is sensitive to the dynamic processes at mid-ocean ridges.
The "Earthquakes, Up Close and Personal" session focused on important issues currently at the forefront of earthquake research. These ranged from the very practical, including a discussion by Tom Heaton on recent earthquake ground motion results and their relevance to building response, and a talk by Art Frankel on the new national hazard maps currently recently released by the US Geological Survey. Frankel also illustrated how a large-scale PASSCAL deployment could make important contributions to a better understanding of regional seismic hazard. Frankel and Heaton both discussed issues, such as "deaggregation" of seismic hazard and building ductility, that may be relatively alien to the seismological community, but that are critical to a successful interface between earthquake scientists and the public. As Kaye Shedlock pointed out in the lively earlier section on education, the public are the policy makers.
The four other talks in the session discussed open research issues in earthquake science. Raul Madariaga presented recent three-dimensional simulations of dynamic rupture in which spontaneous Heaton pulses are generated even in a model with no intrinsic discreteness. An understanding of the rupture characteristics of the rare, largest earthquakes is necessary to quantify potential future ground motions. Dave Jackson illustrated the importance of understanding the long term magnitude distribution in a given region. In particular, the frequency and size of the rare, largest earthquake can have important consequences in the quantification of hazard. Nano Seeber presented results concerning a hypothesis that has been discussed at some length in the earthquake research community: that the static stress changes caused by an earthquake play an important role in triggering aftershocks (and sometimes other large earthquakes). Seeber´s result showing that it is possible to recover much of the 1992 Landers slip distribution, by assuming the aftershocks to have been triggered by Coulomb stress changes and then finding the optimal slip distribution for the mainshock, provides one potentially important confirmation of the hypothesis. In the final talk, Charlie Sammis presented intriguing results from the application of a critical point model to regional seismicity. He showed that increases in regional seismicity prior to large earthquakes seem to be ubiquitous, and that, according to the critical point model, this increase is due not to interaction between earthquakes, but to the earthquake occurrence reflecting the correlation dimension of the regional stress field. If borne out, these results suggest that it might be possible to identify regions in which the stress field is well enough correlated to produce a large earthquake, although still not to predict individual earthquakes in a classical sense.
We wish to thank the 1997 Workshop Committee: Karen Fischer, Brown University; Doug Wiens, Washington University; and Tom Boyd, Colorado School of Mines.
Group photograph courtesy of Michael Hasting Navel Air Weapons Station, China Lake, CA. For copies of this photo, you may contact Michael Hasting at the following:
mike @geowiz.chinalake.navy.mil
