Deep Earth DIALOG

Number 2 September 15, 1988

This is the second issue of the newsletter of SEDI, an IUGG Union Committee to Study the Earth's Deep Interior. Interest in SEDI is continuing to grow; this issue is being mailed to 335 scientists in approximately 25 countries. Requests for additional copies of this issue or for copies of the first issue should be addressed to David Loper, Geophysical Fluid Dynamics Institute, Florida State University, Tallahassee, Florida 32306-3017, U.S.A. Also news items for the third issue should be sent to the sa me address.


Report of Blanes Symposium

(The following is a copy of a report, prepared by Edward Benton, Durk Doornbos and David Loper, to be published in EOS.)

1. Introduction

How should the seismic irregularities observed in the lowermost 100-200 km of the mantle (the D" layer), be partitioned between thermal and chemical heterogeneity, anisotropy, and core-mantle boundary (CMB) topography? How smooth is the core-mantle boun dary? Do mantle convection and the thickness of D" regulate the core geodynamo? How large is the outward temperature fall across the D" layer? What is the temperature within the core? At what depths, or range of depths, do the vigorous up and downwell ings, demanded by dynamo action, occur within the outer core? Are seismically excited internal oscillations of the outer core observationally detectable in surface superconducting gravimeter records?

These constitute an illustrative sample of the questions discussed by almost 200 scientists at the first SEDI symposium entitled "Structure and Dynamics of the Core and Adjacent Mantle," held on the Costa Brava of Spain, at Blanes from 23-25 June, 1988, in conjunction with the 17th International Conference of the Committee on Mathematical Geophysics. SEDI, an acronym for Study of the Earth's Deep Interior, is a new committee of the International Union of Geodesy and Geophysics, created at the Vancouver General Assembly in August, 1987. The following is a report of the SEDI symposium, which consisted of a thirteen oral presentations, forty eight posters, and five open discussion periods during which the posters and general issues were discussed.

2. Core Dynamics

The liquid outer core, believed to be the seat of the dynamo process which sustains the geomagnetic field and creates its secular variations, is of great interest physically, dynamically, thermodynamically, and compositionally.

Compositional buoyancy, released at the inner core-outer core boundary by the selective partial rejection of the light alloying element of the core upon solidification, continues to hold the favored position as the energy source of the geodynamo. Noneth eless, much discussion concentrated upon possible regions or layers of gravitational instability within the outer core, the distribution of the Brunt-Väisälä frequency in the core and so forth. High-pressure studies suggest a stable compositional gradien t within the outer core. On the other hand, the dynamically allowable density differences within an unstably stratified region are far too small to be detected; H. K. Moffatt (University of Cambridge) estimates the fractional density contrast to be only 3 x 10-9. Work on observing core oscillations, through the tiny gravity signals they produce, is still in its infancy, but D. Crossley (McGill University, Montreal) and D. Smylie (York University, Downsview) have secured funding for a Canadian supercond ucting gravimeter to supplement that operated by P. Melchior (Observatoire Royal, Brussels) as well as a few others elsewhere.

The composition and temperature of the core continue to be controversial topics. High pressure experiments, reported by T. Ahrens (California Institute of Technology, Pasadena), reveal that the melting temperature of FeO exceeds that of pure Fe at core pressures. Consequently, the addition of oxygen appears to raise the melting temperature of a Fe-FeO alloy and the density of the liquid exceeds that of the solid. This appears to eliminate oxygen as the light constituent in a liquid core of binary comp osition and strongly constrains the amount of oxygen that can be present in a multicomponent core. The high core temperatures found by T. Ahrens, R. Jeanloz and colleagues, based on shock experiments and static high pressure experiments involving laser h eating, were not in agreement with the much lower temperatures reported by R. Boehler (Max Planck Institut fur Chemie, Mainz), based on electrical heating.

A number of papers reported efforts to produce maps of the fluid motions at the surface of the core, expecting that this will help constrain models of the more deeply buried geodynamo. The first generation of such models were based on the frozen-flux hy pothesis, which ignores ohmic diffusion in the core, and uses only the vertical magnetic field at the CMB together with additional dynamical assumptions such as lack of up and downwelling, geostrophy, or steadiness. These results are not in good agreemen t with each other. Preliminary work on new models which permit magnetic diffusion and utilize horizontal field components was reported by E. R. Benton (University of Colorado, Boulder) and D. Gubbins and D. Lloyd (University of Cambridge, Cambridge). T aking this model a step farther, J. L. LeMouël (Institute de Physique du Globe, Paris) reported on calculations showing that decade fluctuations in the length of day can balance changes in the estimated angular momentum of the core relative to the mantle.

3. Core-Mantle Coupling and Shape

In 1977, A. Dziewonski, B. Hager and R. O'Connell first showed the feasibility of using P wave arrival times for tomographic studies of the lower mantle. Since then, several groups have used seismic tomography to determine the topography and structure o f the core-mantle boundary (CMB). While the amplitude of the travel-time anomalies are in agreement, the inferred CMB structures reported by A. Morelli (Instituto Nazionale di Geofisica, Rome) and by T. Jordan (Massachusetts Institute of Technology) and K. C. Creager (University of Washington, Seattle) are not in agreement. Morelli infers a significant (ca. 15 km) topography of the CMB whereas Jordan and Creager infer structure within the D" layer.

A large amplitude of CMB topography is at odds with other topographic constraints. J. Wahr (University of Colorado, Boulder) noted that the ycomponent of topography is constrained to be no more than a few meters in amplitude. D. Doornbos, T. Hilton and T. Kristensen (University of Oslo) reported that the dispersion of travel-time data of phases involving bottom-side reflections (PKKP, PKKKP, etc.) requires a very smooth reflector. Also, Melchior's gravity data suggests topography of no more than 200 m .

These three independent data sets together suggest that the irregularities sensed by the tomographic studies are within the mantle rather than at the CMB. This interpretation was strengthened by the report by Doornbos et al that the reflector depth infe rred from PKKP is 10-15 km shallower than that inferred from PcP, suggesting the existence of a thin boundary layer at the base of the mantle. Also, R. Haddon (Geological Survey of Canada, Ottawa) noted that precursors to PKP imply ubiquitous and strong lateral variations at the base of the mantle. It would take large horizontal temperature differences, on the order of 1000°K, to produce the observed signals. There was some discussion of whether persistent compositionally distinct regions within the lo wer mantle could produce the signals, but no consensus was reached.

4. Mantle Structure and Dynamics

Models of the structure a few hundred km above the CMB formed another topic of lively discussion during this meeting. The problem has been discussed in the geophysical literature since Lay and Helmberger in 1983 published their interpretation of certain waveform characteristics following SH at distances near 80 deg. Their interpretation was in terms of a velocity discontinuity a few hundred km above the CMB (later modifications were made to allow for lateral variations). The observations themselves ar e not in doubt but, as T. Jordan (Massachusetts Institute of Technology, Cambridge) and independently J. Schweitzer (Institute of Meteorology and Geophysics, Frankfurt) noted, some of the seismological implications of the Lay-Helmberger model have not bee n observed. Two alternative explanations of the observations were advanced: by V. Cormier (University of Connecticut, Storrs) as slab diffraction of S waves and by R. Haddon as diffraction of low-velocity regions in the lower mantle. The model put forw ard by Haddon differs markedly from that by Lay, but both claim to provide a reasonable fit to the observed waveforms, and both can be put into a framework of mantle convection.

 R. Widmer, G. Masters and F. Gilbert (University of California, San Diego), reported on their attempt to improve the existing spherically averaged earth models using spheroidal and toroidal degenerate seismic normal modes with frequencies below 8 mHz. S urprisingly, they were able to fit the data from mantle-sensitive modes quite well, but were unable to fit the core-sensitive modes. This raises the disturbing possibility that the existing core models are incorrect.

 Copies of the program of the first SEDI Symposium are available from Joe Cain, Department of Geology, Florida State University, Tallahassee, FL 32312.


Report of the SEDI Business Meeting

A business meeting of SEDI, chaired by Ned Benton (University of Colorado), was held on June 25, 1988, at Blanes, Spain, and attended by 31 scientists from 8 countries. The following report serves as the official minutes of that meeting.

 The first item on the agenda was a report by Tom Jordan (Massachusetts Institute of Technology) on the current status of and future plans for ISOP: the International Seismic Observation Period. (A full report of this project was given in the first editi on of the DEEP EARTH DIALOG.) A proposal that SEDI officially endorse ISOP was approved by a voice vote. It was similarly agreed that SEDI should establish a working group to coordinate ISOP with other agencies. Tom Jordan was designated as Chairman of this working group. Anyone wishing to be a member of this working group should contact Tom Jordan at the Department of Earth & Planetary Science, Massachusetts Institute of Technology, Cambridge, MA 02139, U.S.A.

 Next, A. W. Green (United States Geological Survey, Denver) described INTERMAGNET: International Real Time Geomagnetic Observatory Network, which is being developed jointly by the Geological Surveys of the United States and Britain. A global network of manned and automated observatories is being assembled that will transmit geomagnetic data in real time via satellite links to two collection points, one in the United States and one in Britain. This will allow a global picture of short-lived magnetic phe nomena to be assembled without delay. A proposal that SEDI officially endorse INTERMAGNET was approved by a voice vote.

 A suggestion by Coerte Voorhies (NASA Goddard Space Flight Center, Greenbelt) that diverse geophysical data types, including magnetic, earth rotation, geoid and seismic, be simultaneously inverted to yield a unified model of the coupled core-mantle syst em was discussed, but no action was taken. The consensus appeared to be that we do not yet know how to weight the various data sets relative to one another.

Ivan Cupal (Geophysical Institute, Prague) spoke of the potential value of SEDI in bringing scientists from western and eastern countries together, and read a list of names of eastern scientists who would be interested in participating in SEDI activities . On a similar note of bringing together diverse groups, attention was turned to the proposal by Tuzo Wilson that a meeting be held bringing together scientists who study deep earth processes with those interested in plate tectonics (see report of Canadi an SEDI activities). This idea was approved by a voice vote.

 The format for the SEDI Symposia, consisting of a few oral presentations, many posters and time for group discussions, was considered next. There was general agreement that the basic format worked well, although there were some problems with integratio n of the posters into the discussions. A number of suggestions were made, including the idea of having selected discussants give brief summaries of small groups of related posters at the beginning of the discussion period. These suggestions will be inco rporated in the format of the second SEDI Symposium. The business meeting ended with a discussion of possible sites for the second SEDI symposium to be held in 1990.


Second International Dynamo Workshop

In October 1979 the Czechoslovakian Academy of Sciences organized at Alsovice an international workshop on Dynamo theory and the generation of the Earth's magnetic field. The sequel to that workshop, on Earth's core boundary and geodynamos, was held at Liblice Castle near Prague from June 27 to July 2, 1988, and attended by 36 scientists from 10 countries. This SEDI Workshop ran simultaneously with a second, and slightly larger, workshop on New trends in geomagnetism - Modern methods and data bases in rock magnetism and paleomagnetism. Participants from each workshop occasionally and informally attended sessions of the other. The two workshops came together in social activities, and in particular for the celebration of the 60th birthday of Dr. V. Buc ha, the director of the Czechoslovakian Institute of Geophysics. Thanks to the tireless organizational efforts of Ivan Cupal, the workshop was very successful. Paul Roberts, who served as the Western coordinator of the workshop has prepared a full repor t. Copies of the report may be obtained from him at Institute of Geophysics and Planetary Physics, University of California at Los Angeles, Los Angeles, CA 90024. The following is a condensation of his report.

 The activities of the workshop fall into seven categories. (1) Fast dynamos are defined as kinematic dynamos whose growth rate is proportional to the strain rate in the limit of large magnetic Reynolds number. Studies of these dynamos serve to elucidat e the inductive processes which can occur in MHD, but are not directly relevant to the geomagnetic dynamo. (2) Mean-field electrodynamics is a consistent method of obtaining a simple set of coupled equations describing the evolution of the magnetic fiel d in which the motions are parameterized by scalar or tensor coefficients. Recent work of the Potsdam and Helsinki groups has concentrated on models in which the a coefficient is a function of the magnetic energy density or the total magnetic energy. Th e aim of these parameterizations is to provide for a feedback mechanism which will allow the field strength to equilibrate to a finite value. (3) Core-mantle interactions and secular variation deal with the structure of the core-mantle boundary and of th e fluid flows and magnetic field at the top of the core. Two current issues are whether core-mantle coupling is principally topographic or magnetic and what information can be gleaned from the electromagnetic signals observed at the surface about the vel ocity field at the top of the core and the underlying dynamo process. (4) The study of core dynamics and the MHD geodynamo has increasingly turned from kinematic dynamos, in which the velocity field is prescribed, to the full MHD problem in which the vel ocity field is one of the unknowns. F. Busse and K. Zhang reported on their efforts to develop a full dynamo model by following a sequence of bifurcations of the thermally driven dynamo equations. It is anticipated that we will see in the coming years s erious efforts to model the geodynamo. (5) An important unresolved question is whether the dynamo is of Taylor type or model-z type. Earlier contentions that the dynamo could be of the weak-field variety, with toroidal field strength comparable to the p oloidal have been discounted. The current debate is whether core-mantle coupling is important, giving a model-z dynamo, or irrelevant, giving a Taylor type dynamo. M. Proctor reported on studies that show the dynamo may be in a state of isorotation in w hich contours of equal toroidal shear and meridional field lines are nearly coincident except near the core-mantle boundary. A number of studies of plane-layer models are being conducted to help elucidate the crucial mechanisms and thus resolve this ques tion. (6) The dynamo mechanism undoubtedly involves waves and instabilities within the Earth's core. This remains a difficult and complicated subject, in which many results appear to be counterintuitive. (7) The energy budget and the long-term behavior of the geodynamo are governed by the thermal interactions between the core and mantle. The favored energy supply remains compositional convection. It was pointed out that this mechanism can be parameterized by a buoyancy source at the bottom of the out er core and a uniform buoyancy sink distributed throughout the outer core.

Projects and National Activities

Canadian SEDI Activities

A Canadian National SEDI group was formed at the IUGG General Assembly in Vancouver in August 1987. Since then meetings have been held at the Fall AGU meeting in San Francisco in December 1987 and at the Canadian Geophysical Union Meeting in Saskatoon i n May 1988. At the last of these meetings, a proposal from Tuzo Wilson was communicated by Doug Smylie, concerning the need to bring surficial geoscience problems together with dynamics of the deep interior. It is generally agreed (among SEDI participan ts) that answers to plate tectonic problems will come only by looking at deeper processes. The Canadians will be investigating the possibility of holding a symposium on this connection.


British SEDI Activities

With the advent of the latest generation of supercomputers, numerical solution of the full nonlinear hydromagnetic dynamo problem for parameter values appropriate to the Earth and other planets has become feasible. At the same time our understanding of many of the basic dynamo mechanisms has reached a level where such a numerical solution is appropriate. Recognizing this, Professor Andrew Soward, Dr. Chris Jones of the University of Newcastle upon Tyne, Dr. Mike Proctor of the University of Cambridge a nd Dr. David Fearn of the University of Glasgow have proposed a project whose ultimate goal is such a numerical geodynamo model. This has recently received funding from the Astronomy and Planetary Science Board of the Science and Engineering Research Cou ncil of the United Kingdom to support two postdocs for up to three years, and to purchase 2 SUN-3 workstations and about 400 hours of Cray X-MP time.

 The goals of the project are twofold. Firstly, to carry out computations to attempt to resolve the difficult question of whether the geodynamo is of the model-z type proposed by Braginsky, i.e., that core-mantle coupling has an important controlling eff ect on the dynamics of the core, or alternatively that the prescription proposed by J. B. Taylor is correct, i.e., that the mantle has no effect in the limit of vanishing viscosity. Secondly, it is planned to develop a time-stepping program to solve the full magnetohydrodynamic dynamo problem with the driving force being derived from unstable entropy gradients in the presence of strong rotation and magnetic field. This is a regime likely to be met in the Earth and leads to motion of global scale.


United States SEDI Activities

The United States SEDI group was formed on December 7, 1987, at the Fall AGU meeting in San Francisco. The 31 people attending this organizational meeting chose Thomas Ahrens (California Institute of Technology, Pasadena) as Chairman and Thorne Lay (Uni versity of Michigan, Ann Arbor) as Secretary. Subsequently, Tom petitioned the AGU to have the group constituted as a Committee of the AGU. This request has recently been approved.

 The AGU Committee for SEDI is charged with the responsibility of providing a focus within AGU for fostering and coordinating activities related to understanding the Earth's deep interior. Included in these activities will be: A. Definition of appropria te research programs ranging from those for individuals to those for large organizations: B. Communications with and among AGU sections; C. Building interactions with other appropriate scientific organizations, both national and international; D. Educa tion of the AGU membership and public about the nature and importance of the problems; E. Organizing of multidisciplinary meetings, workshops, and sessions. The activities of this Committee will be reviewed yearly by its Executive Committee to determine its future role, and will report to the president of AGU. The initial membership of the Executive Committee is T. Ahrens (Chairman), T. Lay (Secretary), M. Drake, R. Hemley, T. Jordan, D. Loper, P. Olson, J. Wahr, and J. Woodhouse.


Call for formation of SEDI national committees

Ned Benton has issued a call for the formation of SEDI national groups in order to facilitate interactions between diverse groups and to better coordinate activities. Anyone interested in forming a national group should contact him at the Department of Astrophysical, Planetary and Atmospheric Sciences, P. O. Box 391, University of Colorado, Boulder CO 80309-0391, U.S.A. It should be noted that this is not a prerequisite for participation in SEDI; individual scientists not affiliated with a national gr oup will still be able to participate in SEDI activities on an equal footing with others.


Plans for future SEDI activities

1. 1988 Fall AGU Meeting, December 5 - 9, 1988

Gerald Schubert (University of California, Los Angeles) will convene the first AGU-SEDI symposium at the 1988 FAll AGU Meeting in San Francisco on "Lower Mantle Dynamics: Plumes, Slabs and the D"-layer, and Interactions with the Core". Anyone wishing to contribute to this symposium should indicate this when submitting their abstract to the AGU and send an extra copy to G. Schubert, Department of Earth and Space Sciences, University of California, Los Angeles, CA 90024.

2. IAGA Assembly at Exeter, July 24 - August 4, 1989

Jean Louis Le Mouël (Institut de Physique du Globe, Paris) has agreed to investigate the possibility of having a SEDI symposium at the Exeter assembly, and to chair the session if it can be accommodated at this late date. Tentative plans are to have a h alf- or full-day session consisting of invited talks by scientists in disciplines related to interests of Division 1, who normally would not attend the IAGA assembly.

 Also it is hoped that we will be able to have a short (half- or one-day) SEDI workshop at Exeter on Saturday, July 29. The purpose of this workshop is to develop a plan of coordinated scientific activities among the various national groups active in SED I. This workshop will be open to all who are interested. If this workshop can be arranged, more specific information will be distributed at a later date.

3. IASPEI Assembly at Istanbul, August 27 - September 1, 1989

SEDI has requested to serve as joint sponsor, with the IASPEI Committee On Physical Properties of the Earth's Interior (T. Yagi, Chairman) and the International Commission for the Lithosphere, of a 2-day symposium on "The earth's deep interior and subduc ting slabs", to be held August 31 - September 1, 1989 The convenors are S. Karato, D. Doornbos and T. Jordan. for further information, contact Professor S. Karato, Department of Geology and Geophysics, University of Minnesota, Minneapolis, MN 55455, U.S .A.

4. Second SEDI Symposium in 1990

There has been much discussion of possible sites, organizers and topics for the second SEDI Symposium to be held in 1990, and how to coordinate this activity with the regularly scheduled conference of the Committee for Mathematical Geophysics (CMG). The CMG traditionally has its conferences in even numbered years the week of the summer solstice, and the first SEDI Symposium was held in conjunction with the 17th CMG Conference at Blanes, Spain, this past June. This arrangement was convenient for those w hose interests span both CMG and SEDI. Consequently there was considerable sentiment for continued close cooperation. On the other hand, the joint meeting was uncomfortably large - approximately 200, making group discussions difficult. Also the sentime nt was expressed that SEDI should hold at least one meeting on its own, in order to establish its identity. After much discussion between the officers of SEDI and CMG at Blanes, it was decided that the CMG will plan to hold its meeting near the summer so lstice in 1990 and inform SEDI of its plans at the earliest possible time. (Anyone interested in the CMG Conference should contact one of the officers: Chairman Albert Tarantola, Institut de Physique du Globe, 4 Place Jussieu, Paris Cedex 75252, France; Vice-Chairman W. R. Peltier, Department of Physics, University of Toronto, Ontario M5S 1A7 Canada; Secretary John Orcutt, Institute of Geophysics and Planetary Physics, Scripps Institution of Oceanography, La Jolla, CA 92093.) SEDI will be free to hold i ts symposium when and where it deems best. An informal invitation has been extended by Ivan Cupal to host the Symposium at Liblice Castle, and a number of other possible sites are under investigation. The choices of site, time and topic for the symposiu m are being coordinated by Ned Benton.

5. IUGG General Assembly in Vienna in 1991

These plans are still quite tentative, but it is likely that there will be a SEDI symposium at the General Assembly. Also, a request has been made to the IUGG that one of the Union lectures be designated the SEDI lecture. The implication is that the sp eaker will be nominated by SEDI, with final approval being given by the organizers of the General Assembly.

Back to SEDI Newsletters