GP52A-01 INVITED
New Data Bases and Standards for Gravity Anomalies
Ever since the use of high-precision gravimeters emerged in the 1950's, gravity surveys have been an important tool for geologic studies. Recent developments that make geologically useful measurements from airborne and satellite platforms, the ready availability of the Global Positioning System that provides precise vertical and horizontal control, improved global data bases, and the increased availability of processing and modeling software have accelerated the use of the gravity method. As a result, efforts are being made to improve the gravity databases publicly available to the geoscience community by expanding their holdings and increasing the accuracy and precision of the data in them. Specifically the North American Gravity Database as well as the individual databases of Canada, Mexico, and the United States are being revised using new formats and standards to improve their coverage, standardization, and accuracy. An important part of this effort is revision of procedures and standards for calculating gravity anomalies taking into account the enhanced computational power available, modern satellite-based positioning technology, improved terrain databases, and increased interest in more accurately defining the different components of gravity anomalies. The most striking revision is the use of one single internationally accepted reference ellipsoid for the horizontal and vertical datums of gravity stations as well as for the computation of the calculated value of theoretical gravity. The new standards hardly impact the interpretation of local anomalies, but do improve regional anomalies in that long wavelength artifacts are removed. Most importantly, such new standards can be consistently applied to gravity database compilations of nations, continents, and even the entire world. Although many types of gravity anomalies have been described, they fall into three main classes. The primary class incorporates planetary effects, which are analytically prescribed, to derive the predicted or modeled gravity, and thus, anomalies of this class are termed planetary. The most primitive version of a gravity anomaly is simply the difference between the value of gravity predicted by the effect of the reference ellipsoid and the observed gravity anomaly. When the height of the gravity station increases, the ellipsoidal gravity anomaly decreases because of the increased distance of measurement from the anomaly- producing masses. The two primary anomalies in geophysics, which are appropriately classified as planetary anomalies, are the Free-air and Bouguer gravity anomalies. They employ models that account for planetary effects on gravity including the topography of the earth. A second class of anomaly, geological anomalies, includes the modeled gravity effect of known or assumed masses leading to the predicted gravity by using geological data such as densities and crustal thickness. The third class of anomaly, filtered anomalies, removes arbitrary gravity effects of largely unknown sources that are empirically or analytically determined from the nature of the gravity anomalies by filtering.
GP52A-02 INVITED
A Preliminary Full Spectrum Magnetic Anomaly Database of the United States With Improved Long Wavelengths for Studying Continental Dynamics
Under an initiative started by Thomas G. Hildenbrand of the U. S. Geological Survey, we have improved the long-wavelength (50-2500 km) content of the regional magnetic anomaly compilation for the conterminous United States by utilizing a nearly homogeneous set of National Uranium Resource Evaluation (NURE) magnetic surveys flown from 1975 to 1981. The surveys were flown in quadrangles of 2° of longitude by 1° of latitude with E-W flight-lines spaced 4.8 to 9.6 km, N-S tie-lines variably spaced, and a nominal terrain clearance of 122 m. Many of the surveys used base-station magnetometers to remove external field variations. NURE surveys were originally processed with IGRF core-field models, which left behind non- uniform residual trends in the data and discontinuities at survey boundaries. In this study, in place of the IGRF/DGRF, we used a spatially and temporally continuous model of the magnetic field known as the Comprehensive Model (CM), which allowed us to avoid discontinuities at survey boundaries. The CM simultaneously models the core magnetic field and long-wavelength ionospheric and magnetospheric fields, along with their induced components in the earth. Because of the availability of base-stations for removing external fields, we removed only the core-derived geomagnetic field based on CM4 (spherical harmonic degree 13) for our compilation. The NURE data have short-wavelength (less than 30 km) noise due to cultural sources, base-station offsets, and residual external field effects. It is possible to reduce and even remove these defects by identifying and editing them and by applying leveling and micro-leveling. There are also many high resolution individual surveys over the U.S. which could be incorporated into the improved NURE database; however, this could take a few years. Therefore, we have created a preliminary full spectrum magnetic anomaly database by combining short-wavelength magnetic anomalies from the North American Magnetic Anomaly Map (NAMAM) and long-wavelength anomalies from NURE using a Gaussian filter centered at 50-km wavelength. We call this product the NURE-NAMAM2008 magnetic database. NURE- NAMAM2008 is useful for analyzing geodynamic aspects of the crustal and mantle magnetic field that require precise long-wavelength information; e.g., estimating Curie-temperature depths and constraining lithospheric temperatures. Preliminary studies show that the corrected long-wavelength components in NURE- NAMAM2008 lead to more realistic Curie depths for the average western U.S. crust.
GP52A-03
The Role and Practice of Property Optimisation to Help Evaluate 3D Geological Models using Gravity and Magnetic Data
As the shift from 2D to 3D geological mapping gathers strength and the number of multi-component potential field data sets increases, there is a need for greater sophistication in the gravity and magnetic modelling tools that can be used to help evaluate and refine the properties and geometry of the various units within these models. The hitherto standard approach of 2D forward modelling of selected cross-sections is progressively giving way to full 3D forward modelling. An example of a user-guided optimisation method to streamline what would otherwise be a time-consuming and frustrating manual iterative refinement process in 3D is presented. A combination of the density and magnetic properties assigned to each geological unit is derived such that the total calculated response best matches the supplied scalar, vector or tensor gravity and magnetic field observations, subject to specified levels of uncertainty (bounds) in the properties. Numerical optimisation is achieved with a standard linear least squares routine, subject to equality and bounds constraints. The user is presented with 3 standard options for every property, allowing the property values to be either (a) fixed, (b) free to vary within a specified range, or (c) free vary over a very broad range. Additionally, properties for a group of geological units can be linked so that they all return the same value. The parameterisation of density properties is relatively straight forward with a single property for each geological unit. Magnetic properties present more of a challenge. Three distinct scenarios are identified and a separate option can be selected for each geological unit. These assume (1) only induced susceptibility, (2) a combination of induced susceptibility and remanent magnetisation of know direction, or (3) a combination of induced susceptibility and remanent magnetisation of unknown direction. In this latter case, a solution is obtained for the total effective magnetisation in the form of 3 mutually perpendicular magnetisation components. The demonstrated approach could be improved by incorporating a user-specified degree of covariance between the density and magnetic properties for each of the geological units. This would upgrade the approach from separate independent optimisation of density and magnetic properties to a more realistic, robust and discriminating joint optimisation process.
GP52A-04 INVITED
Hydrated Mantle Beneath Circum-Pacific Forearcs: An Earthquake-Hazard Application (?) of Global Geophysical Databases
Thomas Hildenbrand was an acknowledged leader in the acquisition, compilation, and interpretation of continent-scale gravity and magnetic anomaly data. It is in this spirit that we offer an interpretation of the World Digital Magnetic Anomaly Map (WDMAM; Hemant et al., 2007), focusing on the signature of mantle hydration and its possible link to earthquake hazards in Circum Pacific subduction zones. The transformation of metabasalt to eclogite in subducting oceanic crust releases large amounts of water into overlying lithosphere, where it potentially hydrates cold upper-mantle peridotite and produces serpentinite. Thermal models indicate that hydrated upper mantle in many subduction zones is cooler than the Curie temperature of magnetite, and, if upper-mantle serpentinite is sufficiently abundant, we should expect long- wavelength magnetic anomalies over subduction forearcs. A crust-mantle model of the Cascadia subduction margin, based on magnetic, gravity, and seismic data, is consistent with the presence of a significant volume of hydrated mantle beneath the Oregon and Washington Coast Ranges (Blakely et al., 2005). Knowledge of hydrated upper mantle has two important applications in assessing earthquake hazards: First, in warm subduction margins, the release of water by slab dehydration is believed to embrittle the descending slab, thereby promoting intraslab earthquakes (e.g., Kirby et al., 1996). Second, the down-dip limit of inter-slab rupture during megathrust earthquakes may be controlled by the position of the hydrated mantle wedge above the subducting slab (Hyndman et al., 1997). New global magnetic databases now allow us to map hydrated mantle in convergent-margin forearcs worldwide. A Fourier analysis of magnetic anomalies from the Circum Pacific, using the WDMAM data, found important magnetic sources at 50 and 160 km depth, which we interpret as upper-mantle sources. Matched filters designed on the basis of these depths were applied to WDMAM data in order to emphasize magnetic fields originating from upper-mantle depths, and a boundary analysis applied to these filtered data yielded a map-based interpretation of forearc hydrated mantle for the entire Circum Pacific. Without exception, all magnetic forearcs seen in our analysis also appear in CHAMP satellite data continued to 350 km (Maus et al., 2007), providing independent evidence of their existence. Our analysis shows evidence for hydrated mantle at many subduction margins of the Circum Pacific, including Cascadia, northeast Japan, the Aleutians, Kamchatcka, southern Mexico, and Central America. All of these subduction margins have thermal characteristics believed to be conducive for intraslab earthquakes. On the other hand, the Nankai, Peru, and southern Chile subduction margins, where intraslab earthquakes also occur, have only minor forearc magnetic anomalies, indicating complexities in the relationship between hydrated mantle and intraslab seismogenesis. In these forearcs, the mantle wedge may contain only small volumes of serpentinite or may be warmer than the Curie temperature of magnetite.
GP52A-05
Potential-Field Data Provide Insights into the Development of Extensional Basins, Aquifer Compartmentalization, and Geothermal Reservoirs--Examples from the Rio Grande Rift, New Mexico and Colorado, and Dixie Valley, Nevada
In the past decade, technological advances in aeromagnetic data acquisition, improvements in 3D modeling of gravity data, and progress in interpretation methods have significantly enhanced the utility of each of these methods for studies of sedimentary basins. Modern aeromagnetic surveys have increased resolution for detecting subtle magnetic contrasts, such as those that arise from shallow basin sediments. In particular, the surveys can be used to map intrabasin faults and reveal a more comprehensive view of fault patterns than can be observed from the surface alone. Modern gravity modeling and inversion methods allow the user to account for variable densities and develop models of basin geometry that are more geologically realistic. Combining interpreted intrabasin fault patterns from aeromagnetic data with 3D gravity models of basin geometry provides additional insights into and constraints on the structural framework of basins and improves applied studies that rely on this knowledge. Several examples of potential-field insights into the development of extensional basins come from the Rio Grande rift, from north-central New Mexico to southern Colorado. These insights show that (1) basins are commonly segmented into smaller sub-basins; (2) rift fault patterns commonly appear to be influenced by pre-existing structures; and (3) local stresses and/or basin tilts changed orientation through time in many places. In examples related to ground-water studies, the aeromagnetic data imply that faults are much more numerous than previously suspected, suggesting that basin aquifers are highly compartmentalized. Gravity models constrain the geometry of the compartments at depth. Potential-field insights of basin structural framework also can be applied to similar extensional basins with geothermal energy resources, such as Dixie Valley, Nevada. In this example, aeromagnetically interpreted faults show a suite of northeasterly trending, curving and branching faults that surround the sides of the deepest part of the basin, as indicated by the gravity data. The producing geothermal reservoir occurs at the north end of this area, where the faults merge together.
GP52A-06
Toward Effective Application of Potential Field Studies to Resource Appraisal: Case Studies From Alaska and the Circum-Arctic
We present examples of the application of gravity and magnetic interpretation to USGS evaluation of undiscovered mineral and hydrocarbon resources in Alaska and the circum-Arctic and highlight the connection of these efforts to the earlier work of Tom Hildenbrand. Effective application of gravity and magnetic data generally involves (1) development of interpretation methods, (2) assembly and processing of potential field data, and (3) application of potential field interpretation constructively to the quantitative assessment of resource potential. Tom influenced and improved all three of these aspects of the science. Much of his interpretive work emphasized the importance of connecting potential field results to real geologic questions. Several examples underscore the continuing challenges in this regard. For example, assessing the possibility of deep hydrocarbon plays within the National Petroleum Reserve, Alaska, requires an understanding and mapping of depth to basement and basement character. This analysis requires the application of Fourier filtering techniques improved by Tom. A key responsibility of USGS scientists is the stewardship of data collected with public money. For example, the assembly and public release of regional magnetic compilations for Alaska and the circum-Arctic are important to fair and open evaluation of hydrocarbon potential and for geologic studies leading to "Law of the Sea" territorial claims. These compilations depend on innovative ways to level regional data developed in part by Tom. Regional geophysical data play an important role in the geologic and tectonic study of frontier areas devoid of other geophysical and geological datasets. For example, interpretation of a prominent regional aeromagnetic high in northern Alaska was influenced by Tom's detailed studies of the rift structure in the New Madrid seismic zone. Considering the legacy of Tom's work it is fair to say that he recognized and participated actively in the major potential field developments that occurred during his career. Tom had a real sense of the promise of new techniques and approaches. By example and active encouragement, he helped guide the careers of colleagues within the USGS and in the global potential field community.
GP52A-07
Recent airborne magnetic data prefer SWEAT reconstruction of Laurentia with Antarctica and Australia to others
Airborne magnetic data provide a means for guiding reconstructions of Rodinia, in particular the hotly debated western continuations of Laurentia, such that the magnetic data tie existing isolated interpretations of geologic units through continuous data coverage, provide plate scale views of geology and tectonics and extend interpretations of units buried beneath cover. Recently released digital continental-scale aeromagnetic compilations, as well as new aeromagnetic data from Antarctica that provide a glimpse of the sub-ice Precambrian geology in critical areas for reconstructions, are useful for plate-tectonic scale reconstructions. Compiling digital magnetic data to map hidden Precambrian basement was a specialty of Tom Hildenbrand, which we apply to global plate reconstructions. We combine magnetic data, plate reconstructions and regional geologic mapping to help constrain the SWEAT (southwest US /East Antarctica), AUSWUS (Australia-Western US) and Sears-Price reconstructions of the Laurentian portion of the 1100-750 Ma Rodinia supercontinent. 'Piercing points' have been used to match Precambrian cratonic blocks and orogenic belts thought to be pieces of the same ancient continent. We identify sources of magnetic anomalies associated with key piercing points in each continent and then match anomalies across continental boundaries within each of the three reconstructions. The only reconstruction in which similar magnetic anomalies can be matched with similar sources in the adjacent continent is the SWEAT reconstruction. Magnetic highs associated with 1.4 Ga A-type granites in the southwestern US correspond to similar magnetic highs in East Antarctica. Although the sources of the magnetic anomalies in Antarctica are buried, a strong signature of 1.4 Ga material in detrital zircon populations along the Antarctic margin and discovery of a glacial clast of A-type granite whose age and geochemical signature match the 1.4 Ga Laurentian granites, suggest that these rocks could be the source of the magnetic anomalies. Sparse magnetic data in Antarctica preclude matching other magnetic anomalies from Laurentia. However, magnetic anomalies associated with 1.85-1.9 Ga magmatic arcs in Canada do not match anomalies in Australia in the SWEAT reconstruction. This suggests that Laurentia was not adjacent to Antarctica and Australia at this time or that they were together but the similar terranes have been faulted away.
GP52A-08
Obtaining Vector Magnetic Field Maps of Geological Samples with Single-Axis Scanning Magnetic Microscopes
Magnetic scanning microscopy can be used to study inhomogeneous magnetization in geological samples with submillimiter spatial resolution. In particular, Superconducting Quantum Interference Device (SQUID) microscopes offer a unique combination of high spatial resolution and outstanding field sensitivity. However, due to physical constraints, most magnetic microscopes only measure a single component of the magnetic field. Nevertheless, Maxwell's equations can be used to demonstrate that the components of a static magnetic field in a region of space devoid of sources are not independent. This means that single-axis scanning magnetometers can potentially obtain all of three components of the field external to the sample. We present an improved technique in the Fourier domain which can obtain the complete vector field planar map from just the planar map of one component. This technique is fast, robust, does not rely on any specific source type or configuration and does not require the formulation of an inverse problem. In contrast to other applications in geomagnetic remote sensing, the assumptions and conditions imposed on the field distribution by the technique can be naturally satisfied in scanning microscopy of geological samples. We analyze the advantages and shortcomings of the technique, and establish which sensor and mapping configurations may yield high quality three-component field maps with virtually no information loss. We present results obtained both with synthetic data and experimental data measured with our SQUID microscope system.