A workshop has been held May15-17, 1997 at IGPP, La Jolla,
to discuss several aspects of a new reference Earth model. We agreed that
the new reference earth model should not just be a 'tweak to PREM'.
We regard this process as a community effort with input necessary from the
wide range of possible users (e.g. mineral physicists and geodynamicists)
and not just seismologists. The workshop format was based on a Plume conference format
with a few (mainly non-seismological) review papers interspersed with sessions
of short representations and discussion. There were also 25 posters to
stimulate discussion outside of the main lecture hall. Review talks were the following:
- S. Karato: The three "an"s and the reference Earth model:
Anelasticity, Anisotropy and Anharmonicity and Earth structure
- D. Weidner: On mantle materials and mineral physics with emphasis on phase
transformations and transition zone
- J. Mitrovica: On flow calculations in the mantle and determination of viscosity
(large-scale structure in the mantle)
- R. Jeanloz: On the core and core/mantle interactions
- G. Nolet: What do seismic waves really resolve in an isotropic 3D Earth?
In a plenary discussion at the end of the workshop, the participants agreed
to recommend the following guide lines.
|
These guide lines are preliminary and open for discussion. Please contact us
( gmasters@ucsd.edu) if you feel that
an important issue has not been addressed.
|
What should we have soon?
- Availability of REM guidelines, datasets, and codes on a REM website
- Recommendations as to the scope of a REM (explicit recommendations as to
what parameters, what parameterization(s), what specific datasets, and
what specific codes should be used in the REM - at least for the 1D part)
This is what this website is all about.
A preliminary version of
Exchange Format
Specifications have been summarized by Pete Davis and is open for discussion.
The Holy Grail: What do we expect from a new REM?
It should be an internally consistent dynamical, chemical, mineralogical model
of the Earth predicting:
- seismology (tomography, anisotropy, attenuation,etc)
- gravity/geoid, post-glaical rebound, etc
- topography of all boundaries, heat flow, magnetic field, etc
How many REMs do we need?
There should be:
- a global 1-D REM
- regional 1-D REMs (ultimately tied in to a global 3-D model)
- a ''regionalized'' and ''a-priori'' model of the top 100/200/300(??)km of the Earth
- a full 3-D global model
Examples of Seismological Guidelines
All REMs should:
- be at least density, elastic moduli, and attenuation (or equivalent)
- specify structure throughout the Earth (from center to surface)
- have a sufficiently flecible representation of attenuation to allow datasets at
different frequencies to be mixed (e.g. an absorption band operator with Q
a function of frequency in the band and with corner frequencies that can be
spatially varying).
- have an associated ''error analysis'' (e.g. corridor of possible models?)
Some Recommendations for Specific Regions
CRUST
- Existing models should be improved using short period surface wave data
- Attenuation should be included
- Perhaps include anisotropy?
UPPER MANTLE AND TRANSITION ZONE
- 1-D models should be transversely isotropic (TI in all solid regions?)
- The only ''global'' discontinuities should be 410 and 660
- Define the global discontinuities by width and shape
LOWER MANTLE
- As close to Adams-Williamson as possible (taking account of attenuation)
- Maybe only isotropic?
- D" should usually be distinguished by a second order discontinuity (at what depth?)
OUTER CORE
- Isotropic(?), homogeneous, and as close to Adams-Williamson as possible
INNER CORE
- Should 1-D model be isoptropic?
- Should ICB be simple first-order discontinuity (or define width and shape)?
- Should a ''simple'' anisotropic model be provided?
|
The next meeting is planned to be held at Harvard University late 1998/
early 1999.
|
top
Go back to
main page
Gabi Laske (
glaske@ucsd.edu)