A4. MAIN FIELD AND CORE PROCESSES1
Appendix A4 was largely developed by the following workshop group: David Loper (Group Leader), J. Bloxham, S. Braginsky, J. Cain, M. Fuller, C. Harrison, R. Langel, R. Merrill, L. Newitt, N. Peddie, J. Quinn, P. Roberts, K. Verosub.
A long-term goal of geophysics is a coherent picture of the structure and dynamics of the solid Earth. The geomagnetic field is a central feature in this quest. Observations of the field are a source of data in some studies, while the understanding of the field and its generation results from other studies. For example, paleomagnetism played a historic role by confirming continental drift and ushering in the birth of plate tectonics by providing an explanation of magnetic striping on the seafloor. On the other hand, understanding the origin and nature of the geomagnetic field is a fundamental component of any coherent picture of the Earth's interior.
The evolution and dynamics of our planet affect all of us, yet the processes and forces involved are hidden from view within the deep interior of the Earth. One of the few tools available to probe that interior is the measurement and interpretation of the geomagnetic field. Three sources contribute to the magnetic field near the Earth: currents in the core, magnetization in the upper lithosphere, and currents outside the Earth and their induced components inside the Earth. The core (or main) field is by far the largest; because it must travel through the mantle, this field yields information both on the region of its generation and on the electrical conductivity of the mantle. In extracting this information, the temporal variation of the field is at least as important as its description at a particular epoch. These time variations occur over periods from months to millennia. Variations in the core's magnetic field that have periods less
Page 135than about a year are believed to be screened from surface observation by the electrically conducting portions of the Earth's mantle.
Although there is certainly additional information that might be gleaned from historical records, only a limited number of observations were made, and only a small fraction of those observations survive. Furthermore, many of the most interesting variations of the geomagnetic field occur on time scales that are greater than those accessible through the historical record. In order to extend the history of the Earth's magnetic field beyond the limit of direct measurements, one must turn to the methods and techniques of archaeomagnetism and paleomagnetism. Many natural materials, such as sediments, lava flows, and baked clays, preserve a record of the geomagnetic field existing at the time of the formation or transformation of the material. With appropriate laboratory measurements and procedures, the paleomagnetic record of the geomagnetic field can provide fundamental constraints on the longer time-scale behavior of the geodynamo.
The variety of time and length scales of interest in the study of the main field and core processes is illustrated in Figure A4-1. This is a double log plot with the time extending from 1 year to 5 × 109 years (that is, the age of the Earth) on the horizontal axis and length extending from 1 m to 6 × 106 m (that is, the radius of the Earth) on the vertical axis. Here the crustal fields are a source of noise (except for those components contributing to paleomagnetic measurements). The processes of interest are enclosed in rectangles and the data sources are enclosed in ovals. It can be seen from this figure that, except for geomagnetic jerks, the phenomena of interest occur over long times (≥ 103 years) and on large scales (≥ 105 m), whereas the data are almost exclusively measured contemporaneously in more recent times or on much smaller scales.
The basic premise that virtually everyone accepts is that the Earth's magnetism is created by a self-sustaining dynamo driven by fluid motions in the Earth's core. As to the energy mechanism for those motions, the majority favors convective driving, most probably of compositional origin; a minority believes that the luni-solar precession is the source. The theory of convective dynamos is a challenging branch of
Page 136magnetohydrodynamics (MHD), one in which progress is continually being made, but only slowly. It is known theoretically that such dynamos can operate in both weak-field regimes and strong-field regimes. Although the theory of weak-field dynamos is easy to understand, at least compared with that of strong-field dynamos, most theoreticians believe at the present time that the geodynamo is of strong-field type, in which Coriolis and magnetic (Lorentz) forces are of comparable magnitude. There is at the present time a considerable divergence of opinion about the form a strong geodynamo would take. It centers on the speed of the geostrophic flow within the core and the strength of core-mantle coupling. More concentrated research initiatives are needed to accelerate the removal of uncertainties and a progression towards consensus.
Whatever the outcome of this debate, one thing is clear: the MHD dynamo can take many forms and, without reference to the observational facts and their interpretation, it will be impossible to discover which of the infinity of possible models most closely resembles the Earth's dynamo. The model sought must display the complicated features of the geomagnetic field (from the large irregular reversals to the short-period secular variation) as well as its more regular periodicities and structure. Ultimately the dynamics and thermodynamics of the mantle will have to be incorporated.
The study of the dynamo and its role in the overall Earth system is rooted in geomagnetic phenomena measured at and above the Earth's surface. These data divide naturally into two sets. The first consists of contemporary and historic observations made by satellites, observatories, and surveys; these data are used to construct models of the main geomagnetic field and its secular variation. The second set consists principally of the imprints of past field configurations in a variety of media including baked artifacts, lavas, lake and sea sediments, and magnetized rocks. This information tells us of the past state of the field.
Page 138Present-Day Secular Variations
The secular variation of the Earth's magnetic field, that is, the temporal variations with periods of a few years and longer, provides one of the few probes of time-dependent processes in the Earth's deep interior that can be used in the attempt to unravel the dynamics of this inaccessible part of the Earth. Most other probes of the Earth's interior, such as seismology, are more limited in that they provide only a snapshot of the interior.
In order to study processes in the deep interior, the field at the Earth's surface must be sampled rather densely, so that the observations can be used to construct maps of the magnetic field at the core-mantle boundary, the upper boundary of the region of greatest interest. The data for the past 300 years or so have been sufficient to carry out this program, but not at high resolution. High-quality results have been possible only recently, with the availability of accurate, almost perfectly spatially distributed, satellite observations. Without doubt, one of the most crucial needs for main field geomagnetism is to ensure that data are gathered from evenly distributed permanent magnetic observatories and/or satellites, and that this continues in the future on a regular basis, ideally providing continuous monitoring of the field.Paleo-Variations
The mere existence of a paleomagnetic record stretching several billion years into the past— implying the existence of fluid motions in the core for at least that long—provides a valuable constraint on the evolution of the Earth. One of the best-documented features of the Earth's magnetic field is that it has often reversed its polarity. However, in the past 10 years, new paleomagnetic studies have provided records of polarity transitions with considerably more detail than was previously available. The new records demonstrate that the behavior of the transitional field is far more complex than was previously believed and that to understand
Page 139even a single polarity transition requires a global distribution of high-resolution records.
Progress is being made in this direction, and high-resolution, multiple records are now available for several recent transitions. Some of these records are from the Southern Hemisphere, which until now has been significantly underrepresented in the data base. Many additional records from both hemispheres are needed. Several of the new records have resulted from closely spaced sampling of cores of marine sediment, demonstrating the potential of these cores for studies of polarity transitions. Back-to-back transitions recorded at a given site sometimes show considerable similarity—indicating that the factors which control the details of the transition process may persist over long periods of time. Thus, it is important to study not only a given transition at many sites but also successive transitions at the same site.
An additional source of information about the field derives from the analysis of geomagnetic excursions—short-term, high-amplitude fluctuations in paleomagnetic directions. Although the existence of many reported geomagnetic excursions is still a matter of discussion, some geomagnetic excursions have enough spatial and temporal consistency that they can be accepted as valid geomagnetic phenomena. This acceptance raises the question: should these excursions be regarded as large-scale secular variations or as aborted polarity transitions? Only with more intense study of the phenomenon of geomagnetic excursions can this question be resolved.
The pertinent scientific issues may be categorized by three questions that are considered in more detail in the following sections:
How can better spatial and temporal descriptions of the main geomagnetic field be obtained?
How should those descriptions of the main geomagnetic field be interpreted in terms of mantle properties and core processes?
How can a theory of core processes be provided?
The description of the existing main geomagnetic field and its secular variation is based on data that are imperfect and often inadequate for the desired purposes, which require extrapolation to the core-mantle boundary. The data contain errors and have uneven coverage in time and space. Furthermore, the contributions from the crust and core are mingled and thus far are inseparable. The core is believed to dominate up to harmonic degree and order 12; the crust dominates above 16; both sources contribute in the interval 13 to 15.
It is important that better methods be devised to optimize the information content of the data used in the description of the field and secular variation and to place confidence levels on the resulting models.
The time history of the geomagnetic field over the past few hundred years can be determined from a combination of data from long-running observatories and ship-track records. While ship-track records from England have been used successfully to model past secular variation, the reliability of these reconstructions could be significantly increased by “mining” the historic records of other maritime nations, such as the Netherlands, Spain, and Portugal.
The longer history of the field, on time scales of 103 to 105 years, is determined by a combination of records from archaeomagnetism, lavas, and lake and sea sediments. These data are crucial in the determination of the spectrum of oscillations of the geomagnetic field and possible variations in the strength of the field. This is an important time scale, as there are theoretical reasons to believe that the fundamental oscillation period of the dynamo operates in this range. Lavas are of use in obtaining detailed time-histories of the geomagnetic field at select intervals of time. Of particular interest are those that measure a reversal transition or a geomagnetic excursion.
The long-term behavior of the geomagnetic field, on time scales of 106 years or more, is determined principally from paleomagnetic measurements. One area of interest is the average field behavior during a given polarity interval. To first approximation, the long-term average field direction at any given site is that of a geocentric axial dipole field. However, general agreement is lacking as to how long it takes for that mean field direction to manifest itself. Indeed, several studies have suggested that the time-averaged field contains nondipole components that persist throughout a given polarity interval. The existence of such components would place a fundamental limitation on the use of paleomagnetic data to determine plate motions and tectonic movements.
Marine magnetic anomalies provide a first-order record of the polarity transitions for the past 200 million years. However, that record contains certain time intervals where major problems exist and certain unusual features whose precise nature has not yet been determined. Collateral studies of contemporaneous rock sequences exposed on land or obtained by coring at sea are needed to resolve these problems.
Because geomagnetic polarity transitions are the only frequent, globally synchronous geophysical phenomena, high-precision magnetostratigraphy is critically important to the correlation of ocean sediments and to the establishment of a temporal framework for biostratigraphic and isotopic events. An accurate magnetic polarity time scale is imperative for any attempt to place regional events in the context of global change. In addition, an accurate time scale is also a prerequisite to analyses of the statistics of the polarity intervals, which provide information directly relevant to dynamo modeling. Among the topics that have been studied and for which definitive answers do not yet exist are the distribution of the lengths of polarity intervals as well as differences between normal and reversed intervals. Recently there has been considerably controversy about the possibility of periodicities in the frequency of reversals—again without a definite resolution.
While additional work on the time scale for the last 200 million years is still needed, perhaps the most important area for new research will be in determining the magnetic polarity time scale for time intervals prior to the past 200 million years. Because there are no marine magnetic
Page 142anomalies and no deep-sea marine sediments of this age, this task will be particularly difficult. However, it is possible that the major questions about the statistics of the polarity intervals cannot be solved without such data. How Should Those Descriptions of the Main Geomagnetic Field Be Interpreted in Terms of Mantle Properties and Core Processes?
The downward extrapolation of contemporary models of the field is complicated by the fact that the mantle has a nonzero electrical conductivity which increases with depth. While the conductivity of the upper portion of the mantle (down to 1,000 km at most) may be estimated from magnetic-induction studies, the conductivity of the lower portion is determined by the character of the geomagnetic secular variation signal that propagates from the core. Variations in this signal shorter than several years appear to be screened by this conductivity, masking from view any rapid variations (≤ 1 year) within the core. However, these same data provide valuable information about this conductivity distribution, particularly that in the lowermost portion of the mantle, called the D” layer. Until now, models of the conductivity distribution have assumed axisymmetry. However, there is strong evidence from seismology that the D” layer is nonsymmetric. An important question is whether the geomagnetic data can discern this nonsymmetric conductivity distribution.
Reliable determination of the temporal evolution of the magnetic field at the core-mantle boundary provides a probe of the dynamics of fluid flow in the core (of obvious importance to dynamo theory) through the determination of the pattern of fluid flow immediately beneath the core surface in much the same way that surface plate motions constrain models of circulation in the mantle. Many important questions need to be answered. Is the flow nearly in geostrophic balance, as in the atmosphere, or is the complicating effect of Lorentz forces important? Is the top of the core stably stratified? What is the radial length scale of the flow? In other words, is the flow imaged at the core surface representa
Page 143tive of whole core convection, or is the flow as it is seen indicative of flow only in the near vicinity of the core-mantle boundary and which may be highly influenced by lateral heterogeneity of the mantle? Do abrupt changes in the magnetic field result from magnetic instabilities, such as kink instabilities, or are they just an ingredient of “normal” secular variation?
A related problem concerns core-mantle interactions. These can be split into two classes: those that transfer angular momentum between the core and mantle and which thus affect the length of day, and those that do not involve, at least directly, the transfer of angular momentum, such as thermal interactions, but which are likely to affect the pattern of fluid flow at the core surface. Great progress has been made in the past few years in understanding the transfer of angular momentum: the budget of the Earth's total angular momentum is now well explained on time scales of decades (meteorological studies have led to a similar understanding on seasonal to weekly time scales). But, although the budget is well understood, the mechanism is not, and the great challenge is to understand the relative importance of the various torques (pressure, electromagnetic, gravitational, and viscous) that operate across the core-mantle boundary. This issue is important not only in an understanding of the dynamics of core-mantle coupling; it also has implications for other fields, including dynamo theory and the interpretation of Earth nutation. Thermal and chemical interactions play a central role in determining the convective motions throughout the Earth's interior and are a vital ingredient in the understanding of the thermal and chemical history of the Earth.
There are several questions regarding the nature of a polarity transition:
Is the Earth's magnetic field predominantly dipolar at the Earth's surface during a transition?
Are there preferred directional systematics present during a transition?
What is the temporal relationship between intensity changes and directional changes?
What similarities and differences are there between successive transitions?
Is the transitional field predominantly of the dipole (asymmetric) or quadrupole (symmetric) family?
Are there very rapid changes in direction and/or intensity during a polarity transition?
For many years, it was thought that westward drift was the dominant mode of behavior of the nondipole field. Recent compilations of historical data suggest that individual centers of nondipole activity can move and evolve in a variety of ways.
What processes govern the behavior of the sources of the nondipole field?
Are there persistent geographical controls on this behavior?
What are the characteristic time constants in the secular-variation record?
Is there a change in secular variation immediately preceding or following a polarity transition?
What is the relationship between geomagnetic excursions and secular variation?
Are reversals distinct from normal secular variation or do they represent an end member of a continuous spectrum in secular variation?
Recently, paleomagnetic secular-variation data from lava flows have been used to provide valuable indirect information on the magnitudes of the competing dynamo families (dipolar and quadrupolar). Is the ratio of the strength of dipole to quadrupole family high during the Paleozoic long reverse interval as it appears to be during the Cretaceous long normal interval? What do differences in the mean normal-polarity and reverse-polarity data reflect? What are the changes in time of the mean field-polarity asymmetries? What other paleomagnetic data can be used to provide information on deep-Earth properties and processes?
The present determinations of the temporal variations in intensity of the field on both short and long time scales are especially significant; key intervals are the past few hundred thousand years and during the superchrons.How Can a Theory of Core Processes Be Provided?
A recent renewal of interest in the precessional driving mechanism was sparked by the discovery that flows with elliptical streamlines—somewhat similar to the flows that are driven by the luni-solar precession—are unstable. Whether this has any significant implications for core dynamics is as yet unclear, but it certainly deserves further investigation.
There is a very great need for further large-scale numerical computation in which MHD dynamos of the same general type as that of the Earth are constructed. Not only would this help resolve the nature of the strong-field balance, but it would also indicate how models can be “tuned” to the geomagnetic data. Ultimately, this tuning will lead to new information about the physical state of the core. Very many numerical experiments need to be performed covering a wide spectrum of input parameters, and the results need to be compared. Moreover, there should be several groups engaged in this research—groups that will benchmark each other's programs, devise increasingly efficient numerical techniques, and generally stimulate each other. So far, no efficient numerical method has been devised that can cope with the dynamics over periods of the order of the free decay time of field (14,000 years) in a system that turns once every day. And of course, because of the celebrated theorem of Cowling, it is necessary to seek fully three-dimensional solutions. It is a little alarming for the standing of the United States in this field that large-scale computational efforts are being made in this country at only one institution. Other nations are making a greater commitment. For example, the United Kingdom has a substantially larger effort in manpower and resources in this area.
Over the past decade, there has been an increasing awareness of the importance of the upper layers of the core and the lowest (D”) layer of the mantle. A better understanding of these regions and of the nature of core motions is emerging from the geomagnetic data gathered by satellite and, over longer periods, by geomagnetic observatories. Clearly, more insights will emerge from further work in this area, particularly regarding core-mantle coupling and the decadal variation in the length of the day.
The source of secular-variation impulses has baffled theoreticians for more than a decade. It remains an important unsolved problem of geomagnetism.
Models and charts of the main field and its secular variation have a wide range of practical applications. They provide quick answers to questions about the strength and direction of the geomagnetic field. Navigators, surveyors, and scientists depend on them directly, and nearly everyone makes use of them indirectly. They are the source of the vital magnetic information shown on the millions of nautical and aeronautical charts and topographic maps sold in the United States each year. Models are built into the onboard navigation systems of countless military, commercial, and private aircraft. They provide the declination needed for radio navigation systems and for airport runway designations. Surveyors often need to know the declination, obtained from a model or chart, for surveying land and mines. Models are used by geophysicists to remove the main field trend from aerial survey measurements taken in the search for minerals and petroleum. Information on declination, obtained from models and charts, is used for aiming antennas and drill strings. Models are used for finding the paths of cosmic rays, for the magnetospheric coordinate system, and for calculating field-line geometry and the locations of conjugate points. Magnetic charts are especially useful for visualizing the shape of the Earth's magnetic field.
Accurate knowledge of the present-day magnetic field at the surface is important to a wide range of commercial and military activities,
Page 147including natural resource exploration, navigation, and orientation of drill holes in oil exploration and production.
The reversal chronology and secular-variation records provide valuable global and regional dating tools for diverse purposes. They can be used to help establish baselines in global change research or for stratigraphic control in mineral exploration.
A severe difficulty in modeling the present field is the extremely uneven distribution of observatories over the Earth's surface and the absence of continuous monitoring by satellite. To obtain information about the longer time scales of the secular variation, a major effort in archaeomagnetism and paleomagnetism is required, including the utilization of new techniques of dating and magnetic measurement. The theory of core dynamics requires greater support in manpower and computing facilities.Improved Global Models of the Geomagnetic Field
The accuracy of any main field model, both in time and space, depends critically upon the quality and distribution of the data upon which it is based. In the absence of global, vector satellite survey data, modelers are dependent on the various types of surface data. Even if periodic satellite surveys were available, determination of accurate temporal change between surveys depends upon the surface data. For study of the Earth's interior, determination of the temporal change is at least as important as determination of the field itself, and the time scales range from less than a year to centuries; therefore, the highest priority in data needs is for temporally continuous and global vector data. The best-quality data would be from Magsat-like satellite surveys.
In the absence of such surveys, the highest priority should be to augment the present distribution of magnetic observatories to obtain as
Page 148nearly as possible an equal area coverage over the entire Earth, with a spacing of no more than 2,000 km. If, in the future, continuous satellite survey data become available, the augmented observatory data network will still play a vital role in modeling. This is because truly accurate models will be those which also model the ionospheric field. Since satellite data are acquired above the ionospheric currents, such fields are a source of inaccuracy for field models based only on satellite data. The combination of satellite data with a good distribution of surface data permits separation of the measured field according to its three constituent fields: the interior of the Earth, the ionosphere, and the magnetosphere.
For incorporation of a definitive model of the ionosphere into our main field models, data from more than a single satellite are required. This is because the morphology of the ionospheric fields varies in local time, whereas the data from a satellite are acquired at only two local times (one ascending, the other descending). A reasonable attempt at such a definitive model could probably be carried out with three satellites in orbit simultaneously and spaced equally in local time.
The network of observatories, even if augmented as described above, can profitably be supplemented by true repeat stations. By that is meant stations that are visited every 2 to 5 years for remeasurement of the field. The measurements at each reoccupation should be taken at the physically identical position, and care should be taken to minimize or eliminate any magnetic disturbance fields and the daily variation.
The observatory data should be supplemented by periodic surveys over oceanic and remote land areas by aircraft or ship. These, in fact, are crucial in any areas where the observatory data do not meet the requirements stated above. Such measurements will never have the accuracy or continuity so valuable in observatory data, but—in the absence of an observatory—will at least prevent the models from being wildly inaccurate.
Because of the importance of determining the temporal characteristics of the geomagnetic field, there is great interest in studying past field changes. This can be done only for the few hundred years during which some sort of magnetic data were acquired. Recent efforts to collect such historical data and use them in deriving spherical harmonic models have
Page 149proved very successful and useful. However, large portions of the historical data are not known or available to the modeling community. An effort to gather such data is needed to increase the accuracy of our historical models. Long-Period Variations: The Need for Archaeomagnetic and Paleomagnetic Syntheses
With modern and historical data, the longest temporal scales accessible are between 60 and 400 years. These are not the only important time scales for study of the geomagnetic field and the underlying dynamo. In fact, it can be argued that longer-period variations are more important and fundamental.
Information about the longer-term secular variation of the geomagnetic field can be obtained from the paleomagnetic study of archaeological materials, lava flows, rapidly deposited sediments, and cave deposits. One clear advantage of using archaeological materials is that they can often by placed in a fairly restricted chronological context. The accuracy and resolution of that context is greatest for samples that are only a few hundred years old. Moreover, samples of that age are an important means of supplementing the historical record of field directions. In addition, paleointensity studies can provide the intensity information that is lacking from the historical observations.
Lava flows represent another source of information about geomagnetic secular variation. The advantage of lava flows is that they are accurate recorders of the geomagnetic field direction at the time of the eruption of the lava. In addition, using relatively well-established and well-understood procedures, it is possible to determine fairly accurately the absolute intensity of the geomagnetic field at the time of the extrusion of a lava flow. However, here too, rock magnetic studies are needed to improve the existing techniques. In addition, the record of volcanic activity in some fields extends over periods of several hundred thousand to a few million years. Paleomagnetic study of these fields can provide
Page 150information about the long-term behavior of the field that is difficult to obtain from any other source.
In contrast to the archaeological materials and lava flows, rapidly deposited sediments have the potential of providing continuous records of geomagnetic field behavior. Recently paleomagnetists have succeeded in obtaining reproducible records of secular variation from lacustrine environments. For at least the past 10,000 years, these records are consistent on a regional scale and contain features that can be correlated on a global scale. At the present time, the resolution that can be obtained for any given area is limited by the understanding of the processes involved in the magnetization of a sediment; considerably more research is needed in this area.
The use of sediments in determining changes in the intensity of the geomagnetic field lags far behind their use to determine the direction of the field; very few paleointensity records are available. This discrepancy is caused primarily by the fact that the methodology for extracting paleointensity data from lake sediments is still evolving; there is clearly a need for more fundamental work in this area. Resolution of the existing questions regarding paleointensity techniques would pave the way for determination of the complete paleomagnetic vector. One or more records of the secular variation of the total geomagnetic field would probably provide important new insights about secular variation itself and about the interaction between the dipole and nondipole components of the field. Furthermore, the intensity of the geomagnetic field and the solar wind are the primary modulators of the flux of galactic cosmic rays that control the production of radiogenic isotopes in the Earth's atmosphere. New data about variations in geomagnetic intensity, combined with existing data on radiocarbon production rates, will lead to a better understanding of major solar processes.
Although initially studies of both directional records and paleointensity records have focused on sediments spanning the last 10,000 years, records extending back in time for the last several hundred thousand years are also very important. Extant and dry lakes in unglaciated areas and long cores of marine sediments will be the primary sources of these records.
No single source of information will provide a complete record of secular variation, and in the final analysis, data from several sources will be needed. It is important to recognize that virtually all aspects of paleomagnetism are involved in this process. For example, the results of laboratory rock-magnetic studies are essential to the interpretation of paleomagnetic samples from the field, because we need to know the conditions under which the recorder of the field is reliable.
As these data are accumulated, they will extend knowledge of the behavior of the Earth's magnetic field from the time scale of a few hundred years (available from observatory data) and historical records to a time scale of tens of thousands of years (initially) and hundreds of thousands of years (eventually).Priorities
It is generally difficult to set firm priorities. However, on the basis of the discussion above, the following are suggested as the top priorities, grouped in order of importance:
Augmentation of the observatory network;
Continuous satellite survey; and
Archaeomagnetic and paleomagnetic data acquisition.
Aeromagnetic and shipborne magnetic surveys;
Repeat data; and
Multiple satellite survey.
Programmatic AspectsScientific Organizations
The study of the geodynamo is a central component of SEDI (Studies of the Earth's Deep Interior), an international program under the guidance of a IUGG committee of the International Union of Geodesy and Geophysics (IUGG). This committee organizes biennial interdisciplinary symposia allowing intimate interaction between geomagneticians and scientists from other geophysical disciplines having a common interest in the structure and dynamics of the Earth's deep interior. SEDI also encourages the formation of national groups. Two such groups have been formed to focus on geomagnetic studies: in the United Kingdom, a National Environmental Research Council (NERC)-sponsored cooperative study of hydromagnetic oscillations, and in Japan, a project on The Earth's Central Core.
Under the SEDI umbrella, a Cooperative Study of the Earth's Deep Interior is being developed in collaboration with the Earth Sciences Section of NSF. Among other things, this initiative calls for a sustained cooperative effort in the dynamic modeling of the dynamo process. If implemented, this initiative should result in a significant level of support for this activity.Governmental Agencies
Governmental agencies are involved in two ways: several agencies require the output of modeling efforts, and data acquisition capability crosses agency boundaries.
Modeling requirements exist within DOD, USGS, NOAA, NASA, and NSF. For the latter two, the requirements stem from the use of models by researchers under the grant system.
An important fact is that different data sets are collected under the auspices of different agencies. For example, satellite data are acquired by DOD and NASA; magnetic observatory data are acquired by USGS and
Page 153corresponding agencies abroad; and aeromagnetic data are acquired by DOD, USGS, and the natural resource industry. Each of these organizational efforts is directed toward meeting internal requirements, yet each data set is of value for geomagnetic studies and in particular for modeling the main field. To our knowledge, no formal coordination occurs between agencies.
The initiation of minimal joint planning between agencies, with the participation of a scientific advisory group, would provide guidance so that the data acquisition efforts could be planned both to satisfy needs of individual agencies and to better meet the needs of the modeling community.
Conclusions and Recommendations
Recent advances in geomagnetic analysis and theory, the advent of data from recent measurements onboard satellites and historic measurements on wooden ships, and complementary developments in other geophysical disciplines (for example, seismic tomography) have resulted in new insights into the existing picture of the structure and dynamics of the Earth. These insights have also revealed shortcomings both in that picture and in present attempts to improve it. There are several developments in measurement acquisition and research that are particularly crucial for the continued progress of research in this area. They include the following:
Geomagnetic observatories should be set up to give more uniform coverage over the Earth's surface, especially the oceanic areas. Costs could be minimized by sharing existing facilities (for example, Incorporated Research Institutions for Seismology [IRIS]). On a longer term, there should be a commitment to fly a magnetic satellite at all times.
Archaeomagnetic and lake- and sea-sediment data should be gathered in order to add significantly to the sparse data available about the geomagnetic field during the past few thousand years.
Page 154High-quality data should be sought that cover paleosecular variation, including paleointensity, magnetic stratigraphy, and reversal transitions. The geomagnetic opportunities afforded by the Ocean Drilling Program (ODP) have not yet been fully realized. Rescue archaeology has the potential to provide useful archaeomagnetic information.
Better field models should be generated from the data to provide constraints on geodynamo theory.
When compared with the advances made in the gathering and interpretation of data, the development a model of the geodynamo is seen to be lagging. Future progress would be enhanced if this imbalance were redressed by the commitment of more manpower and advanced computer resources to the modeling of the geomagnetic dynamo.