T51C-0460 0800h
Rheological Consequences of Incipient Melting in Crustal Rocks
A review and reinterpretation of older experimental data on the deformation of partially-melted granite reveals a non-linear strength decrease with increasing melt fractions. This decrease is characterised by two sharp discontinuities, each reflecting a dramatic change of strength within a limited range of melt fractions. A first discontinuity is shown by all experiments at melt fractions of approximately 0.07. The strength drop at melt fractions smaller than this discontinuity is the largest over the entire melting range. Hence the greatest weakening occurs well below the well known rheologically critical melt percentage (RCMP). In contrast to previous interpretations, the RCMP is inferred to occur, at melt fractions of 0.4 to 0.6, for crystallising as well as for melting rocks. However, the magnitude of the stress drop at the RCMP is negligible compared to the stress drop at melt fractions $<$ 0.07. Increasing melt content in the range of melt fractions between 0.0 and 0.07 induces a rapid increase in the percentage of melt-bearing grain boundaries. This increase is not linear and barely existent at melt fractions greater than 0.1. Therefore, the increasing proportion of wetted grain boundaries is suggested to be the primary cause for the dramatic strength drop described above. Extrapolation of the existing deformational experiments on partially-melted crustal rocks to natural conditions remains speculative due to the lack of experiments performed at controlled melt pressure, under conditions favouring steady state, in the dislocation or diffusion creep regime. However, the existing data suggest that the attainment of a melt fraction $<$ 0.07 will control the large-scale localisation of deformation into partially-melted crustal layers, irrespective of the attainment of the RCMP. Hence, very small amounts of melt, eventually too small to be imaged by geophysical methods, may exert a drastic control on large-scale localization of deformation.
T51C-0461 0800h
Crossing the Bridge From First-Order but as yet Qualitative Geophysical Observations to Quantitative Crustal Models
Typically, when geophysical data sets are gathered, quantities are measured e.g. seismic travel times or amplitudes, which can be modelled to obtain information on the material properties e.g. P- and S-wave velocities and the seismic quality factor, Q. However, in many cases other auxiliary observations are made for which as yet no suitable methods have been derived to obtain information about Earth structure. For example, in northeast Tibet, explosion seismology provides good quality record sections for both P- and S-waves. However, to the south in the northern Lhasa and southern Qiangtang blocks in central Tibet, explosion seismology only provides good quality P-waves. S-waves are not recorded as they are too weak. Here, upper crustal S-waves (e.g. Sg phase) can only be recorded using earthquakes and S-wave reflections from the Moho (SmS phase) can only be recorded using those earthquakes with magnitudes greater than 4.5. Further south in the southern Lhasa block, explosion seismology only provides reasonable quality upper crustal P-waves (e.g. Pg phase) and has difficulty in generating P-wave reflections from the Moho (PmP phase). Yet further south in the Tethyan Himalaya, explosion seismology again provides well recorded PmP waves. Similar patterns of observations can also be recognized in the central Andes of northern Chile and in other regions of young tectonic and/or magmatic activity e.g. Kenya rift, Eifel volcanic region. These observational differences clearly correspond to real differences in some element of crustal structure; most probably variations of Q, heat flow, fluid and/or magma content etc. However, up until now there have been no attempts to actually try and quantitatively model these first-order but nevertheless qualitative observations. This contribution will present examples of the above mentioned observations from the Andes and Tibet and will attempt to discuss possible modelling strategies to quantitatively interpret the above mentioned observations.
T51C-0462 0800h
Broadband Seismic Imaging of Western Himalaya, Ladakh and Karakoram
We present crust and mantle structure of the western Himalaya, Ladakh and Karakoram using teleseismic waveform data recorded on a 17 station broadband seismic network. 15 of the stations were operated along a N-S profile from Delhi (29 $\deg$N) to Karakoram (35 $\deg$N) and two on the eastern Ladakh. These stations were operated during Sept. 2002 - Dec. 2003 with station spacing varying from 40-70 km. The station configuration included CMG3T seismometer and REFTEK 72A-07 data logger recording waveforms at 20 samples per sec in continuous mode. We modeled good quality earthquake waveforms in terms of crustal S-wave velocity variation with depth and also the mantle layering using P-to-S converted receiver function approach. Important results are as follows- The Moho depth increases progressively from 40 km beneath Indo-Gangetic plain to $\sim$60 km below the Zanskar. Further north, below Indus suture, Ladakh and Karakoram the Moho depth is 70, 75 and 80 km, respectively. Apart from the mid crust low velocity below Indus suture and Ladakh, we observe a very significant velocity inversion in the mid crust from Indo-Gangetic plain to Main Central Thrust. The 410 and 660 km discontinuities are sharp and generally agree with the global reference model IASP91 along the profile. We also modeled the P-wave travel time data to define lateral velocity variation in crust and upper mantle.
T51C-0463 0800h
Crustal Deformation and Crust-Mantle Interaction in Active Collision Zones: Central Tibet and East Anatolia
A major question in the tectonics of collisional mountain belts concerns the relative coupling or decoupling of the crust and mantle. In this respect, recent IRIS PASSCAL seismic experiments in the Tibetan and East Anatolian plateaus provide an important opportunity to study crustal deformation and crust-mantle interactions in continental collisions. Receiver functions (RF) from the dense INDEPTH III seismic array that was deployed across central Tibet, show a crustal thickness of $\sim65$ km and a very weak Moho beneath the Bangong-Nujiang suture (BNS) that could be due either to a gradational velocity contrast or scattering by topography on the Moho. By using a global minimization technique, we modeled the azimuth dependant variations on RFs recorded at a representative station within the suture and found strong anisotropy near the surface and in the middle crust separated by south dipping ($\sim25\deg$) layer. Mid-crustal anisotropy occurs in a low velocity zone (LVZ) and has a fabric dipping gently ($\sim18\deg$) northward that might be related to a well-developed near-horizontal rock fabric induced by crustal flow. On the other hand, a preliminary analysis of data recorded by the Eastern Turkey Seismic Experiment (ETSE) show a drastic change in crustal structure between the Arabian plate and East Anatolian plateau across the Bitlis suture. RFs show a strong Moho ($\sim40$ km) and a mid-crustal discontinuity ($\sim25$ km) beneath the Arabian plate and a slightly weaker Moho ($\sim45$ km), a mid-crustal LVZ ($\sim25$ km) and a mantle discontinuity ($\sim85$ km) beneath the East Anatolian plateau. Arrival times of multiples also indicate low Vp/Vs ($\sim1.70$) for the Arabian plate and high Vp/Vs ($\sim1.80$) for the plateau. In the central region of the plateau, the fast directions of SKS splitting and Pn anisotropy are parallel to each other but at high angles to the GPS motions suggesting crustal flow within the LVZ that decouples surface deformation from the upper mantle. In contrast, for the Arabian plate, the GPS motions are parallel to the Pn but different from the SKS fast directions and may reflect a coupled crust and upper mantle. In order to test this idea, we will model crustal anisotropy that is characterized by systematic tangential energy and polarity reversals in the existing data and interpret its tectonic significance.
T51C-0464 0800h
Crust and upper mantle structure of the Southern Basin and Range and Colorado Plateau of Arizona: Project COARSE
Project COARSE (COnsortium for Arizona Reconnaissance Seismic Experiment) has operated a temporary broadband seismic network across Eastern Arizona for the past 18 months. COARSE stations sample lithosphere across the Southern Basin and Range and onto the Colorado Plateau. Here we present observations aimed at improving our understanding of the creation and maintenance of the stable Colorado Plateau and highly deformed Basin and Range, which has undergone substantial compression and extension. Forces related to subhorizontal subduction of the Farallon plate beneath the North American plate followed by a slab steepening or detachment event at ~35 Ma are common to geologic and tectonic models of the deformation and volcanism within the western United States. However, how these tectonic events have left their signature in the lithospheric structure remains unknown. Receiver functions from COARSE stations display distinct structural differences between the Colorado Plateau-Transitional Zone and Basin and Range stations including variations in crustal structure, Moho signature, low velocity zones within the crust and upper mantle, and crustal anisotropy. A clear high amplitude Moho arrival corresponding to crustal thicknesses near 30 km can be seen in receiver functions from the southern Basin and Range. Differently, receiver functions from Colorado Plateau stations exhibit a high amplitude, negatively polarized arrival at ~1-2 seconds which could result from a layer of upper crust containing anomalously slow shear wave speeds. Additionally, the Ps converted phase from the Moho has considerably lower amplitude than those from the Basin and Range possibly indicative of reduced impedance contrast across the Moho, or that reverberating energy from shallow structures obscures the primary Moho arrival. To investigate these possibilities, we have begun to construct a suite of models that can produce the observed receiver functions from the Colorado Plateau and Basin and Range. Utilizing waveform modeling to create synthetic receiver functions from models of Vp, Vs, density, and crustal thickness, we attempt to determine what combination of characteristics best match our observations at each station. Preliminary modeling results indicate that a 30 percent drop in Vs centered about a 3.25 km thick low velocity body at 9 km depth matches observed receiver functions for the USNSN station WUAZ on the plateau. We speculate that the low velocity feature may relate to the presence of partial melt beneath the Plateau that is associated with Pleistocene age volcanism in the vicinity.
T51C-0465 0800h
Seismic Properties and Acoustic Emission of Olivine and Molten Basalt: A Laboratory Study
Here, we present the first results of petrophysical investigations on the system olivine + MORB up to 1473K and 300MPa confining pressure using an internally-heated Paterson gas apparatus especially designed for the measurements of physical properties of rocks. Both Vp and acoustic emissions were measured during heating and cooling of cold pressed synthetic aggregate of both San Carlos olivine and MORB powder (in the latter case till complete melting). The same measurements were also made on a layered sample comprising an olivine-MORB-olivine sandwich. Vp was measured using the pulse transmission technique, with 1 MHz piezoelectric transducers mounted on the end of ceramic buffer rods. The buffer rods were calibrated using a single crystal of sapphire cut parallel to the c-axis. The acoustic emissions were recorded using the same transducers. A significant decrease of both velocity and of the amplitude of the seismic signal were observed during experiments involving MORB at temperatures above 1273 K. BSE imaged that melt from the MORB layer had intruded into the olivine matrix, even under hydrostatic conditions. Higher acoustic activity recorded after the melting onset, characterized by short durations and very low amplitudes, confirms the melt migration, and suggests a brittle intrusive mechanism such as hydrofracturing. Importantly, this provides a unique experimental insight into magma migration in the lithosphere and the mechanism of dyke emplacement. The results support the presence of a low but distinguishable seismic output during this process.