Studies of the Earth's Mantle from Long-Period Ground-Based and Satellite EM Data Chairpersons: N. Grammatica and J. Pous 3.1 SATELLITE MAGNETIC ANOMALIES AND INDUCTION ANOMALIES FOR VARIOUS HETEROGENEOUSLY CONDUCTING EARTH MODELS - A COMPARATIVE STUDY Naphsica Grammatica and Pascal Tarits IUEM - UBO, UMR 6538 "Domaines Oceaniques", Place Nicolas Copernic, 29280 Plouzane, France naphsica@sdt.univ-brest.fr Using the solar quiet daily variation as a source field, we solve the three-dimensional induction problem in three types of heterogeneously conducting earth models. The first model consists of an upper shell with a laterally varying electrical conductivity representing the continents and oceans. The underlying mantle is assumed to have a radially varying electrical conductivity. The second earth model includes subduction slabs in the upper mantle in addition to the land-ocean distribution in the uppermost shell and in the third model, conducting zones are added in the upper mantle below the Pacific Ocean floor. Using a statistical non-parametric test, we observe, in selected regions, a correlation between our calculated magnetic anomalies and anomalies observed on a global Magsat magnetic anomaly map. We show that the long-wavelength magnetic signature of subducting slabs, at satellite altitude, tends to either reinforce or attenuate the coast effect, depending on the respective electrical conductivities of the subducting slabs and the surrounding upper mantle. Finally, we show that the long-wavelength electromagnetically induced anomaly field in the degree range 5- 27, contributes a magnetic signature at satellite altitude that can reach up to 6 nT at dusk local time. 3.2 INVESTIGATION OF SATELLITE MAGNETIC DATA FOR INDUCTION STUDIES: MAGSAT, OERSTED Pascal Tarits IUEM, UMR 6538, Place Nicolas Copernic, F-29280 Plouzane, France tarits@univ-brest.fr The time varying magnetic field recorded at satellite altitude is the result of magnetospheric (external with respect to the satellite) and ionospheric (internal with respect to the satellite) electromagnetic activity and of the internal induced response of the conductive earth. Investigation of the vectorial MAGSAT data showed that global transfer function between the inducing magnetospheric field and the induced internal field could be derived from satellite magnetic data. Estimates of the spherical symmetric conductivity structure of the earth mantle as well as large scale asymmetric features have been derived from the data and compared to the mantle conductivity obtained from land-based observatory stations. A step forward in satellite data analysis for induction studies from space is to model simultaneously the earth mantle conductivity and the source field in time and space domain in order to take advantage of the spatial resolution inherent to full coverage of the magnetic field provided by the satellite. The approach is reviewed and necessary hypotheses discussed. A preliminary analysis of the OERSTED data is presented. 3.3 A SPHERICAL HARMONIC ANALYSIS OF THE DST RECOVERY PHASE OF MAGNETIC STORMS Ulrich Schmucker Planckstrasse 19, D-37073 Goettingen, Germany uschmuc@uni-geophys.gwdg.de The analysis is carried out separately for the Dst recovery phase of individual magnetic storms, using hourly means, and the Dst continuum over many storms, using low-pass filtered decimated hourly means. Daily variations are removed either by subtracting mean daily variations from the analysed storm-time section, or by low-pass filters with a cut-off at 0.75 cpd for the continuum. Observatory data from the IGY/C (1957-59) and the IQSY (1964- 65) are used. By excluding high-latitude locations a total of 70 to 90 observatories with simultaneous records are available. The spherical harmonic analysis of individual storms is conducted hour by hour after storm begin, both for the horizontal and vertical components. The resulting potential coefficients are split into external and internal parts. EM response estimates are derived from them either in the time or frequency domain. For the Dst continuum the order of analyses is reversed. Twelve spherical harmonics are used. It is found that two days after storm-begin they have all disappeared except for the ringcurrent term. To account for deviations from a spherically symmetric Earth estimates are bi- variate for the respective term and also the ringcurrent term as the principal source of induction. Any anomalous portion in the ringcurrent term remains undetectable, however. 3.4 EARTH TRANSIENT ELECTROMAGNETIC RESPONSE AT SATELLITE ALTITUDES TO GEOMAGNETIC STORM TIME RECOVERY PHASE Mark E Everett Department of Geology and Geophysics, Texas AM University, USA colt45@beerfrdg.tamu.edu A numerical simulation is presented of transient electromagnetic induction in a spherical 3D Earth by an external, time varying symmetric ring current whose intensity and decay behavior typify the recovery phase of a large geomagnetic storm. The electrical conductivity distribution of the Earth is characterized by an inhomogeneous mantle underlying a surface heterogeneous shell whose conductance reflects the geographic distribution of oceans, shelves, continental interiors, coastal plains and sedimentary basins. Calculations are made in geomagnetic coordinates. The transient electromagnetic response is evaluated at mid-latitudes along a twilight, sun-synchronous orbit to indicate the magnitude of spatiotemporal variations that would be recorded by a space-based magnetometer platform, assuming that core, lithosphere and ionospheric contributions have been removed. Methods are investigated for applying the simulation results to fully 3D spatiotemporal analysis of actual satellite induction data. The implications of the research for probing lateral variations in upper mantle electrical conductivity are discussed. 3.5 LINKAGES BETWEEN EXTENSION, PETROLOGY, AND MAGNETOTELLURICS IN THE CRUST AND MANTLE: AN EXAMPLE FROM THE CALIFORNIA BASIN AND RANGE Stephen K. Park Institute of Geophysics and Planetary Physics, University of California, Riverside, CA 92521 USA magneto@ucrmt.ucr.edu A wideband (0.01-20000 s) magnetotelluric survey across the southern Sierra Nevada and the California Basin and Range has identified zones of enhanced conductivity in the lower crust and upper mantle that underly the resistive batholithic rocks at depths greater than 10-20 km and beneath some parts of the extended terrain to the east. These conductive zones extend to depths in excess of 200 km, based on model sensitivities, and lie well below the Moho depth of 32-38 km from seismological studies. Therefore the enhanced conductivity in the mantle cannot be attributed to conventional explanations for conductive continental crust and is instead likely due to partial melt. Such an interpretation is consistent with gravitational, seismological, and geologic evidence. Estimates of the partial melt fraction range from 2 to 5% at depths of 40-70 km, but this high an estimate poses problems with the amount of basalt and with residence times in the mantle. Petrological observations of accessory phases in mantle xenoliths may help resolve this dilemma. Small amounts of highly conductive sulfides are found in peridotitic xenoliths from late Quaternary basalts. Equilibration temperatures from the xenoliths are sufficiently high that the sulfides likely coexist in a molten state with the basaltic melt. Sulfides are extremely conductive relative to the solid matrix or the basaltic melt, so a small fraction can increase the bulk conductivity of the mantle appreciably. Previous estimates of 2-5% partial melt based on inferences of partial melt from MT measurements can be plausibly reduced to less than 1%. Such low melt percentages have longer residence times in the mantle and are more consistent with the volumetrically minor late Quaternary basalt flows and the primitive basalt compositions. 3.6 THE ELECTRIC LITHOSPHERE OF THE SLAVE CRATON Alan G. Jones(1), Ian J. Ferguson(2), Alan D. Chave(3), Rob Evans(3), and Gary W. McNeice(4) (1) Geological Survey of Canada, 615 Booth St., Ottawa, Ontario, K1A 0E9, Canada (2) Department of Geological Sciences, University of Manitoba, Winnipeg Manitoba, R3T 2N2, Canada (3) Woods Hole Oceanographic Institution, 360 Woods Hole Road, MS 8, Woods Hole, MA 02543-1539, U.S.A. (4) Geosystem-Canada Inc., Pickering, Ontario, Canada ajones@cg.nrcan.gc.ca The Archean Slave craton, in the northwestern part of the Canadian Shield, has become an international focus of broad geoscientific investigation following the discovery of diamondiferous kimberlite pipes in the early 1990s and the opening of North America's first commercial diamond mine in October 1998. As part of LITHOPROBE's efforts to understand the history and tectonic development along the SNORCLE transect, three electromagnetic surveys, using the magnetotelluric, MT, technique, have been carried out on the craton since 1996, of which two involved novel acquisition on the frozen lake ice and on the lake bottoms. MT responses sampling 50-150 km depths exhibit a maximum spatially located with the Eocene kimberlite field in the Lac de Gras region. Resistivity models show a spatially-confined anomalous upper mantle conducting region, of resistivity less than 30 ohm.m, at depths of 80-120 km beneath the Lac de Gras region. The depth and spatial position of this anomalous region coincides with a geochemically-defined harzburgitic ultradepleted layer interpreted as oceanic or arc-related lithosphere emplaced during early tectonism. Speculative interpretation of the conducting region is in terms of either ionic conduction from hydrogen diffusion or electronic conduction due to the presence of carbon in graphite form, neither of which can be excluded based on the existing evidence. The crust and uppermost mantle above the conductor are more heterogeneous than other regions of the Slave, suggestive of disruption during kimberlite emplacement. 3.7 DEEP LAKE BOTTOM MAGNETOTELLURIC SOUNDING IN THE SLAVE CRATON A.D. Chave(1), R.L. Evans(1), A.G. Jones(2), and J.H. Filloux(3) (1) Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA (2) Geological Survey of Canada, Ottawa, Ontario, Canada (3) Scripps Institution of Oceanography, La Jolla, CA 92093 alan@whoi.edu Deep magnetotelluric sounding beneath Archaean regions can offer new insight into the nature and extent of the cratonic root. However, year-long time series are required to achieve statistically useful results. This poses significant logistical problems at high latitude sites such as much of the Canadian shield or where interference from humans or animals is a concern. A solution to these and other limitations of conventional measurement strategies can be achieved by adapting seafloor technology to work in lakes. Seafloor instrumentation is autonomous and requires very little power, so that long deployments in the secure and stable environment at the bottom of lakes is quite feasible. Ten shallow water EM instruments were deployed in lakes around the Slave province of northwestern Canada for one year beginning in August 1998. The instruments were turned around and redeployed in summer 1999, and are scheduled for final recovery just prior to the workshop. Deployment and recovery was achieved using twin Otter float airplanes staged from Yellowknife at the rate of 3-4 instruments per day. The instruments measure the horizontal electric and three component magnetic fields along with tilt variations and pressure fluctuations. The instruments were programmed to record at a 2.8 s interval for the initial month and 28 s for the remainder of each year-long deployment. This paper will report on the experience achieved using this technique, especially deployment and recovery methods, data quality, and the ensuing response functions. Interpretation of the data in a regional context will be reported in a second paper by Jones et al. 3.8 THE THREE-DIMENSIONAL ELECTROMAGNETIC RESPONSE OF THE EARTH TO RING CURRENT AND AURORAL OVAL EXCITATION Ikuko Fujii(1) and Adam Schultz(2) (1) Kakioka Magnetic Observatory, Kakioka 595, Yasato-machi, Niiharigun, Ibaraki, 315-0116, Japan (2) Department of Earth Sciences, University of Cambridge, Downing Street, Cambridge CB2 3EQ, United Kingdom ikuko@eri.u-tokyo.ac.jp We report on a new global catalogue of the geoelectromagnetic response of the Earth in order to offer an insight on the electrical conductivity of the Earth's mantle. The quality and quantity of the new response catalogue is greatly improved mainly because of four reasons. First, a new database of the geomagnetic hourly means at 225 ground-based observatories from 1957 to 1995 was constructed. Second, a statistically robust procedure using Empirical Orthogonal Function analysis was developed to estimate spatially and temporally coherent geomagnetic field variations in a global scale. Third, we revealed a new source excitation system at 5-107 days which consists of a equatorial ring current in the magnetosphere and two conjugate ring currents at the aurora oval in the ionosphere. The influence of the auroral current system is seen down to geomagnetic latitudes 40 degrees. This implies the existence of non negligible bias on the c response at higher latitudes due to the auroral current system which may have caused difficulties to model the mantle with a 1D layered earth in previous studies. Finally, we propose a new response function d which represents the degree of heterogeneity. Geographical distributions of the c and d responses clearly suggest the existence of large scale heterogeneities of the mantle. 3.9 LONG PERIOD ARRAY MAGNETOTELLURIC MEASUREMENTS AND 3D MODELLING: INDICATION OF AN ANISOTROPIC UPPER MANTLE? Alexander Gatzemeier and Karsten Bahr Institute of Geophysics, University of Goettingen, Herzberger Landstr. 180, 37075 Goettingen, Germany agatzem@gwdg.de In the years 1997-1999 new measurements of the electric and the magnetic field at 31 sites in Central Germany were carried out. Together with old data measured in the last two decades, the array now consits of more than 61 sites. To examine the conductivity structure of the upper mantle beneath Central Germany the magnetotelluric soundings span an area of approximately 300 km x 400 km. With period ranges of almost all sites between 10 s - 10000 s, the data is able to resolve features from the middle and lower crust down to the upper mantle - at least 100 km depth. With a remarkably uniform character at periods above 1000 s nearly all stations show a characteristic splitting of the phase curves up to more than 40 degree with a uniform phase sensitive rotation angle. With 3D forward modelling we try to respond to the question whether the upper mantle beneath Central Germany is electrically anisotropic. To make predications about the type of conduction in the upper mantle, we compare our results with observations of seismic SKS shear wave splitting, which is interpreted as an indication of seismic anisotropy in the upper mantle. 3.10 OCEAN EFFECT IS A MAJOR CONTRIBUTOR TO ANOMALOUS BEHAVIOUR OF COASTAL C-RESPONSES IN DST PERIOD RANGE UP TO 20 DAYS A.V. Kuvshinov(1), D.B. Avdeev(1), O.V. Pankratov(1), and N. Olsen(2) (1) Geoelectromagnetic Research Institute, Russian Academy of Sciences, 142092 Troitsk, Moscow Region, Russia (2) Danish Space Research Institute, DK-2100, Copenhagen, Denmark a.kuvshinov@mtu-net.ru The ocean effect in local geomagnetic C-responses is considered to be small for periods greater than a few days (e.g. Kuvshinov et al., 1990; Takeda, 1992; Tarits, 1994; Weiss and Everett, 1998). We revise this inference by detailed and systematical model studies in the period range from 2 to 64 days, with subsequent comparison of the modelled and experimental (Schmucker, 1978; Roberts, 1983; Schultz and Larsen, 1987; Olsen, 1998) C- responses. The models include a radially symmetric (1-D) section that is overlain by a spherical surface shell. The shell conductance is compiled using 5'x5' NOAA ETOPO topographic/bathymetric and 1x1 degree^2 sediment thickness (Laske and Masters, 1997) maps. The modellings were performed for shell spatial resolutions of 5x5, 2x2 and 1x1 degree^2 and for various underlying 1-D sections. The source is described by a first spherical harmonic, in geomagnetic coordinate system. Comparisons were conducted for coastal geomagnetic observatories Apia, Gnangara, Hermanus, Kakioka, Kanoya, and Simosato where anomalous behaviour of the local C-responses has been previously detected (cf. Schultz and Larsen, 1987). Our first conclusion inferred from the comparisons is that World Ocean is nevertheless major contributor for peculiarities that observed in coastal C-responses in period range from 2 to 20 days. Another conclusion is that to reproduce the experimental data while modelling, one needs to exploit the shell spatial resolution of 2x2 degree^2, or denser, and to include a high- resistive lithosphere in 1-D section.