Transcript Tomographic Imaging of the Crust and Upper Mantle in the

Tomographic Imaging of the
Crust and Upper Mantle in the
Southern Great Basin Interpretation
Glenn Biasi
University of Nevada Reno
Seismological Laboratory
Method and Source Data
Inversion: SIRT, with back-azimuth
binning and model updates damped
inversely to the hit count.
Fit explains 75% of the model RMS (0.23
sec.); model slowness rms: 0.95 sec/km.
90 stations, >7700 travel-time delays
• Data – Relative
teleseismic travel-time
delays
• Invert by projecting
delays on the ray path
• Preferred regional
model 450x450x300
km – 15x15 km blocks
of variable thickness.
• Local model: 4.5x4.5
km blocks
Station Coverage
Note station coverage north, east of
the TM/SC Caldera Complex
90 stations total
Portable stations supplement
analog and digital coverage
around Yucca Mountain
Velocity Scaling
Relationships
• Factors: Temperature, composition (degree of
depletion), melt fraction, and degree of hydration
• UM near solidus – temperature variations likely are small
< 100 degrees C, <1-2% dVp
• Composition: A small effect. (Schutt and Lesher, 2005) dVp <
0.7% for depletion < 20% (cpx out); to 2.5% at 40%.
• Velocity very sensitive to retained melt fraction F
(Hammond and Humphreys, 2000): –3.6% dlnVp/dF
• Hydration
– Subsolidus: OH in UM decreases velocity, shear strength
– Hypersolidus: water escapes to melt -> velocity increase at low
retained melt fractions. Volcanism dries, cools, source -> farther
from solidus. Melt retention decrease velocities.
Resolution – Point anomaly
Reconstruction
Blocks at 50-70 km reconstructed at ~30%; 40% at 90-120 km
Map view (NS, EW) resolution excellent.
70-90 km
• Slight depression
in velocity evident
connecting Crater
Flat to Thirsty
Mountain.
• 2-5 Ma centers
concentrated in 0
dV ring.
Shallow Mantle
YM is on the velocity contrast
<5 Ma basalts concentrate at 1/2% to 1/2% dVp
50-70 km: TM high velocities become
clearer; structure divides the Death ValleyPancake Range lineation.
NE low velocity trend clear, south half
amagmatic
Regional Tomographic
Interpretations
• TM/SC caldera complex caps a depleted hydrous melt
ascent column.
• 2-3% high velocities due to dehydration, cooling, melt
loss.
• Basaltic volcanism follows the margins of the larger root.
Model: residual volatiles +/- heat trigger volcanism.
Waning volumes reflect depletion of volatiles (e.g.,
Crater Flat).
• Large NE-trending low-velocity interpreted as hydrated,
warm, sub-solidus.
– Melt (>1%) should have some surface expression
– Low melt fractions and dry would not explain spatial correlation
with shattered zone in Miocene tuffs and Paleozoic rocks.
Detailed Model
• 4.5x4.5 km blocks, 80 km total model
depth.
• Station corrections from refraction velocity
models.
• Shallow regional model is consistent at
similar wavelengths.
• Interpret for spatial patterns; locally
amplitudes may be excessive.
Local Model
Red line bounds areas with crossing ray
coverage.
CAF: base of the Calico Hills; LC: Lathrop
Cone; LSM: Little Skull Mntn
Velocity amplitude plot clipped at
+-3% - in crust variation can be
silica content
Local Model, 20-45 km depths
-- The ESF is underlain by higher than
background velocity crust.
-- Resolvable structure separates ESF
from North CF cone.
-- NNE trend through Crater Flat – UM
weak zone
-- Shallow mantle depth low velocities S,
SE of LSM may be crustal in origin.
-- Suggested Moho topography
east of the Bare Mountain fault.
Local Model, 45-80 km
Source depth for Crater Flat, Lathrop
inferred at 50-60 km at velocity
increase.
Shallow mantle not homogeneous
North-South Profiles, West of YM
Source areas suggested at 50-60 km
Melt, water extraction leaves above
background velocities
- High velocity crust, upper mantle under YM
-Geometry suggests intrusion or metamorphic
-grade increase, Timber Mntn source
-Low vel. in S. half of Jackass Flats
-Anomaly B source not clear.
Interpretations and
Conclusions - Root
• P-wave tomography provides the third dimension – the upper mantle
today – with useful resolution.
• Root structure separates volcanism north and south of Timber
Mountain.
• Hydrous root may explain unique regional isotopic characteristics
and extensive metasomatism.
– Not a simple subduction enrichment
– Provides hydrous alternative to extreme melting temperatures of Wang,
Smith, et al. (2002).
– Apparent ancient lithospheric Nd, Sr signature may originate with
enrichment source – perhaps water collected on the 410.
• Dry, strong root may protect crust from extension, extensional
faulting
• Near-vertical alignment precludes post-12 Ma (post-15 Ma?) crustal
displacement relative to mantle
Interpretations and Conclusions
– Location of Basaltic Centers
• Quaternary volcanic centers overlie edges of TM/SC
U.M. structure – Not spatially random.
• Outward diffusion, possible upward mobility of water
leads to hydrous “halo”.
• Melting occurs locally where additional water lowers
melting point.
– Melt volume intrinsically limited by water incompatibility and loss
with melt.
– Appears compatible with polybaric melting hypothesis - water in
the garnet lherzolite field is lost to small melt fractions but picks
up garnet melting signature. Shallower melting increases FeO
and SiO2.
• Image suggests a shallow source for the basalts of
Buckboard Mesa. Petrology may support this inference.
Melting Depth and Mantle
Heterogeneity
• Melting depth appears imaged in Crater Flat, Yucca Mountain areas.
– Melt, water removal increases velocities
– Should be compared with petrologic estimates
– May be directly testable with Vp/Vs ratios
• Deep crust and shallow mantle are not homogeneous beneath
Yucca Mountain or Crater Flat.
– PVHA impact on spatial probabilities of future activity.
– Shallow high velocity under Yucca Mountain
– No special low velocities in Crater Flat (c.f., Evans interpretation).
• NNW tectonic trend through W Crater Flat follows mantle low Vp
lineation from Crater Flat to Thirsty Mountain
• Lowest velocity upper mantle interpreted as warm, subsolidus,
partially water saturated.
– Low velocity extends to eastern edge of images.
– West half is amagmatic, north and east are covered with shattered
Miocene and older rocks. Apparently no Quaternary basalt.
Ray Angles in the Upper
Mantle
• Core phases
provide lateral
constraint
• Near tele’s
constrain depth
• Rays are traced
in model but
angles change
little in the U.M.
• Most rays at 300
km sample 100
to 170 km from
the station
Composition
• Schutt and Humphreys (2005):
– spinel lherzolite (shallowest mantle) dVp <
0.7% until cpx exhausted, can reach 2.5% for
40% melt extraction.
– garnet lherzolite (>2-3 Gpa) dVp < 0.5% for
depletion to 40%
– dVp/dVs > 1.8 is diagnostic of melt presence.
• Result: little of dVp is due to modal
change from depletion.
Velocity Sensitivity to Melt
and Hydration
•
Hammond and Humphreys (JGR, 2000) – velocity decreases sharply with
increasing melt fraction in realistic geometries
– 1% melt, 3.6% dVp, 7.9% dVs Dry melting only.
– Melt interconnects at small fractions (<1%, maybe 0.5% or less)
– Lowest dVp < ~4% - suggests a 1% upper bound on melt fraction in area imaged
by upper mantle tomography
– H&H model: 2% melt, ~10% decrease in velocity for (T-Ts)=10 degrees C
– dln(Vp)/dln(Vs) (ratio of percent dV’s) diagnostic of melt presence.
•
Karato and Jung (1998) velocity increases when small fraction partial melt is
present
–
–
–
–
–
Water incompatible, partitions into melt
Water in melt = water not in crystals
Water-free crystal matrix faster as a result.
Limited to low melt fractions
Connectivity at low melt fraction promotes melt loss; residuum higher velocity
after melt removal.
– Could explain neutral velocities under modern caldera (Long Valley)
120-150 km
• Timber Mountain root
is well defined
• Low-velocity region is
wider and lower
amplitude.
150-180 km
• Little changed
from 120-150
km layer.
180-210 km depth
• Root is 1-2%
above average
velocity; 2-3% with
approximate
amplitude
correction
• Root is ~30 km
north and 0-15 km
east of surface
edifice – no strong
crustal shear.
210-250 km depth
• At 210-250 km the
root is ~95 km EW by
105 km NS.
• Shallowest maximum
depth to the base of
the root is over 200
km.
• Slight north plunge
from Moho to this
layer.
250-300 km depth
• NE plunge begins
300-350 km depth
• High velocity body
remains coherent
• Amplitudes max at
base of model
indicate that the true
source could be
deeper.
350-400 km
• Amplitude of NE
structure and
lack of this body
in realistic
synthetic models
indicates high
velocities are
real, if illresolved.
Squeezing: Estimating
Required Model Depth
• Method: decrease model
depth until misfit suffers and/or
imaging artifacts or unphysical
amplitudes result
• Amplitude vs. model depth:
deeper => smaller average
block amplitude.
• RMS misfit vs. model depth:
deeper models fit better
• Spatial distributions of
structure unaffected by model
depth.
Squeezing models