Guided Seismic Waves: Possible MantlePlume Diagnostics

1
Guided Seismic Waves Possible Mantle-Plume
Diagnostics

• Bruce R. Julian
• John R. Evans
• U. S. Geological Survey
• Menlo Park, California

2
Problem Seismological methods offer the highest
resolution mantle images, but current techniques
are

• Most sensitive to large structures,
• Most effective in the upper mantle, and
• Limited by uneven data distribution.

3
Receiver Functions

• Across Snake River Plain near Yellowstone
• (Ken Dueker, pers. comm.)
• Sensitive to discontinuities (upper mantle only)

4
Teleseismic Tomography (of Iceland)

• Resolution is limited to depth lt aperture of
  seismometer array.

Looking South
Looking West
5
Whole-Mantle Tomography

• Model S20RTS (Ritsema et al., 1999, 2004)
• Section along MAR through Iceland
• Resolution best in upper mantle (surface waves).
• Limited by ray distribution, esp. in lower
  mantle.
• Artifacts

6
Finite-Frequency Tomography

• Travel time feels Fresnel zone, of width .
  (Banana).
• Sensitivity 0 on ray! (Doughnut).

7
Lower-Mantle Anomalies (Princeton)

• Based on high-frequency (using ray theory) and
  low-frequency (using finite-frequency theory)
  data.

8
Ray Distribution (Bolton Masters, 2001)

• Plume-like anomalies in SW Pacific correspond
  closely to clumps in data distribution (turning
  points).
• Tomography is limited by uneven data coverage
  much more than by finite-frequency effects.

9
Multiple ScS Core Reflections

• Hawaii earthquake of 1973 April 26, recorded on
  Oahu (Best et al., 1974)
• Relative times indicate high wave speeds and low
  attenuation.

10
ScS Sensitivity Kernel

• Using Banana-Doughnut theory of Dahlen et al.
  (2000)

11
ScS2 Sensitivity Kernel

• Similar to ScS kernel in upper mantle.

12
ScS2-ScS Sensitivity Kernel

• Almost zero small-scale sensitivity in upper
  mantle

13
And Now for Something Completely Different!

• A channel of low seismic-wave speed will act as a
  waveguide Waves cannot escape from such a
  structure, and will follow the channel even
  around (not too sharp) corners.
• Same principal as fiber-optic cable, SOFAR
  acoustic channel in the ocean, etc.
• Observation and identification of such guided
  waves would be virtually conclusive evidence for
  a continuous low-wave-speed channel.

14
Analog Fault-Zone Waveguide

• Theoretical computations from Li Leary (1990)

15
Cylindrical Rod in Homogeneous Medium

• Torsional modes

16
Excitation of Plume-Guided Waves

• Earthquakes in the deep mantle These would be
  great, but they dont occur (we think).
• Teleseisms recorded by seismometers at hot spots
  Adequate?
• Earthquakes at hot-spots recorded
  teleseismically These would produce signals
  comparable to the case above.
• ???

17
A Possible ExperimentTeleseism Recorded by
Seismometer at a Hot Spot

• Caustic (large wave amplitude) for PKP near 120º.
• Various other core phases have similar caustics.

18
Another Possible ExperimentEarthquake at a Hot
Spot

• Exchange source and observer Same result
  (reciprocity principle).

19
Conclusions

• Guided seismic waves are promising tools for
  detecting mantle plumes.
• Positive result would be nearly conclusive.
• Negative result would be ambiguous Absence of
  plume or inadequate excitation of guided wave?
  Interpretation would require detailed theoretical
  computation of excitation by various processes.