V31E-01 INVITED
Imaging the Mount St. Helens Magmatic Systems using Magnetotellurics
A detailed magnetotelluric survey of Mount St. Helens shows that a conduit like zone of high electrical conductivity beneath the volcano is connected to a larger zone of high conductivity at 15 km depth that extends eastward to Mount Adams. We interpret this zone to be a region of connected melt that acts as the reservoir for the silicic magma being extruded at the time of the magnetotelluric survey. This interpretation is consistent with a mid-crustal origin for the silicic component of the Mount St. Helens' magmas and provides an elegant explanation for a previously unexplained feature of the seismicity observed at the time of the catastrophic eruption in 1980. This zone of high mid-crustal conductivity extends northwards to near Mount Rainier suggesting a single region of connected melt comparable in size to the largest silicic volcanic systems known.
V31E-02
Electromagnetic imaging of the deep magma sources of the Taupo Volcanic Zone, New Zealand
The Taupo Volcanic Zone (TVZ), in the North Island, New Zealand is a continental back arc rift associated with the subduction of the Pacific Plate under the Australian Plate and is characterised by the eruption of large volumes of rhyolitic magma during the last 1.6 Ma and an exceptionally high present-day heat flow. Magnetotelluric data had been collected in the TVZ since 1997 to study the deep structure and volcanism of the TVZ but also for exploration of the geothermal systems. We present results from over 200 magnetotelluric soundings covering the central (rhyolitic) part of the TVZ. The data were analysed using 3D inverse resistivity modelling and phase tensor visualisation techniques. Modelling results compare well with the thickness of conductive volcaniclastic material in filling the rift basin and calderas and expected from observed gravity anomalies. The inverse modelling results show a zone of high conductivity in the lower crust and upper-mantle along the central rift-axis that correlates with a zone of high phase observed at long periods. An unusual feature of the MT data at periods of 3-30s is the large phase tensor skew angle values that coincide with the margins of a localized gravity high in the centre of the survey area. This feature appears to be caused by the interaction of a thick near surface layer of high conductive volcaniclastic material with conductive structures at greater depth.
V31E-03 INVITED
Deep crust magmatism at Colima volcano
We present preliminary results from a large seismic investigation centered on Colima Volcano to shed geophysical light on the deep crustal portions of volcanic arc magmatic system. In particularly, we seek to understand the degree to which the specific location of arc volcanoes is governed by anomalous crustal features in contrast to the pattern of mantle sourcing above the subducting slab. We choose Colima because it is long lived, yet geographically isolated from neighboring volcanoes, suggesting either a close correspondence between the volcano and its mantle source, or a well-define crustal feature conducive to volcanism. Colima erupts relatively consistent andesitic compositions suggesting a well-developed "steady state" system. Early results are based on earthquake locations, tomography and receiver functions. Earthquake hypocenters provide a clear image of the subducting slab at approximately 100 km beneath Colima, in close agreement with arc systems worldwide. The slab, as imaged by receiver functions, is in general agreement with the earthquake Benioff zone. The receiver functions also map a well-defined seismic Moho across the region. This requires a clear transition in physical properties between the crust and mantle lithosphere. Though the Moho is unambiguous beneath the volcano, it is perturbed with a lateral scale of tens of kilometers raising the possibility of a focused mantle feature that sources the volcano and that is, perhaps, a primary control on its existence. Early tomographic results show a significant low velocity zone in the lower crust, broadly located beneath the volcanic edifice. We are currently attempting to address the relative roles of magmatic intrusion, crystal fractionation and pre-existing plutonic features in creating the observed lower crustal anomalies.
V31E-04 INVITED
Imaging shallow magma chambers at Alaskan volcanoes with ambient seismic noise
Ambient noise tomography/inversion (ANT) is
an emerging technique in seismology with the
ability to provide 3D images of subsurface
volcanic structure using relatively sparse
seismic networks. The method relies on the
principle that the cross-correlation of noise
recordings at two different seismic stations
reproduces an experiment in which one of the
stations acts as an active source. Ambient
seismic noise in the frequency band from 0.1
to 1 Hz is mostly composed of fundamental
mode surface waves, of both Love and Rayleigh
type. As a result, noise cross-correlations
are sensitive to shear-wave structure and
complement compressional-wave images
computed from phase arrivals of local earthquakes.
At Okmok volcano in the Aleutian islands, a
3D image constructed from 40 days of noise
recordings in 2005 on a 12 station network clearly
shows two low velocity zones (LVZs) centered
about the 10-km-wide caldera: a shallow zone
in the upper 1-2 km and a deeper zone between
4-4.5 km. The shallow LVZ is interpreted to be
weak, poorly-consolidated material within the
caldera; the deeper LVZ is indicative of the
shallow magma chamber at Okmok. That the chamber
is imaged as an LVZ in 2005 points to it remaining
in a molten state throughout the time period
between the 1997 and 2008 eruptions. The existence
of a shallow chamber at Okmok is consistent with
independent studies based on GPS, InSAR, and
petrologic data.
A 3D image has also been determined for the Katmai
group of volcanoes along the Alaska peninsula from
60 days of continuous recordings in 2005 and 2006.
An LVZ at Katmai Pass, previously known from local
earthquake tomography (LET), is evident in the
3D shear-wave velocity model at depths down to 2 km BSL.
That the LVZ exists in compressional-wave velocity
models suggests it is a shallow magma storage
area for Trident volcano. In contrast, low shear-wave
velocity under Martin volcano is likely fluid-related,
given the lack of low compressional-wave velocities in
images derived from LET. Interestingly, a deep (> 5 km),
subtle LVZ imaged between Trident and Mount Katmai
may represent remnants of the magmatic conduit system
from the cataclysmic 1912 eruption of Novarupta.
A deployment of 11 temporary broadband seismometers
are currently in place around Katmai Pass and should
provide more constraints on the structure of the deep LVZ.
The availability of many three-component seismometers
within the Katmai permanent/temporary network makes it
possible to additionally invert Love waves and the
ratio of the horizontal-to-vertical motion of Rayleigh
waves, the HV ratio, to further delineate volcanic
structure from the ambient seismic field.
class="ab'>
UR: http://www.avo.alaska.edu/about/staff.php?view=Seismology&mode=research&dirid=145
V31E-05 INVITED
Dike-induced grabens: InSAR observations and modeling of recent rifting episodes in Afar, Ethiopia.
The 300-km wide Afar depression is located at the junction between the Red Sea, Gulf of Aden and East African rifts. In September and October of 2005, a series of earthquakes and a volcanic eruption signaled the intrusion of a dike along the Dabbahu magmatic segment (in the Red Sea arm) of the Afar rift. Results of elastic modeling constrained by InSAR data (Wright et al, 2006) suggest that the 60-km long segment opened by up to 8 m, between depths of 2-9 km. We investigate deformation mechanisms responsible for the localized zone of subsidence above the emplaced dike, as seen in the geodetic data. Possible candidates include triggered slip on bounding normal faults ahead of the laterally propagating blade-like dike (Rubin and Pollard, 1988), or a distributed failure and collapse of a wedge-shaped block above the dike. We present InSAR data covering four years of rifting activity between 2005-2009 that show continued deformation and repeated intrusions in Afar following the Dabbahu rifting event. We use 3-D finite element models to predict the subsidence associated with the diking event, assuming a brittle/elastic upper crust with a Mohr-Coulomb bulk yielding.
V31E-06
Combining geodetic and petrologic observations to improve constraints on magma chamber dimensions, and magma dynamics
The magma chamber size and constraints on the primary volatile content of intruding magma are determined at Okmok volcano using deformation and petrologic observations along with thermodynamic models that describe the volatiles in solution in the basaltic-andesite melt. Periods of deflation of the volcano following large inflation pulses, recorded by a continuous GPS network, may be caused by degassing of the recently intruded magma. We test this hypothesis with the VolatileCalc solution model and account for magma intrusion, compressibility and volatile degassing effects on surface deformation observations. Magma compressibility is an important factor to consider, because deformation observations are only sensitive to the pressure exerted on or the volume change associated with the chamber walls and compressibility relates these parameters to the mass of an intrusion. The intrusion mass is directly related to the mass of volatiles which, through degassing, later causes deflation of the volcano. In order for degassing to fully explain the deflation signal, the magma chamber below Okmok needs to be on the order of 2-4km in diameter and the intruding magma must have a low CO2 content, <∼500ppm. Although there are not strong constraints on the volatile content of magma at Okmok, the magma chamber size determined here is consistent with tomographic images of a low velocity zone beneath Okmok. This analysis could be improved with more detailed melt inclusion analysis and a detailed record of gas flux during and following an intrusion event.