G44A-01 INVITED 16:00h
Finding Active Faults in a Glaciated and Forested Landscape: the Southern Whidbey Island Fault, Washington
The Puget Lowland, Washington, lies within the Cascadia forearc and is underlain by at least six seismically active and regionally significant crustal faults that together accommodate several mm/yr of net north-south shortening. The surface expression of pre-15-ka slip on Puget Lowland faults has been largely scoured away or covered by glacial deposits, and younger fault geomorphology is often concealed by vegetation and urban development. High-resolution aeromagnetic and lidar surveys, followed by geologic site investigations, have identified and confirmed late Holocene deformation on each of these mostly concealed but potentially hazardous faults. Most geomorphic features identified in lidar data are closely associated with linear magnetic anomalies that reflect the underlying basement structure of the fault and help map its full extent. The southern Whidbey Island fault (SWIF) is a case in point. The northwest-striking SWIF was mapped previously using borehole data and potential-field anomalies on Whidbey Island and marine seismic-reflection surveys beneath surrounding waterways. Gravity inversions and aeromagnetic mapping suggest that the SWIF extends at least 50 km southeast, from Vancouver Island to the Washington mainland, and transitions along its length from northeast-side-down beneath Puget Sound to northeast-side-up on the mainland. Abrupt subsidence at a coastal marsh on south-central Whidbey Island suggests that the SWIF experienced a $M_W$ 6.5 to 7.0 earthquake about 3 ka. Southeast of Whidbey Island, a hypothesized southeastward projection of the SWIF makes landfall between the cities of Seattle and Everett. Linear, northwest-striking magnetic anomalies in this mainland region do coincide with this hypothesized projection, are low in amplitude, and are best illuminated in residual magnetic fields. The most prominent of the residual magnetic anomalies extends at least 16 km, lies approximately on strike with the SWIF on Whidbey Island, and passes within about 27 km of downtown Everett. Glacial deposits are slightly magnetic, and, in places, the magnetic anomaly is associated with topographic lineaments. Spectral analysis indicates, however, that the source of the anomaly extends to depths greater than 2 km and into Eocene sedimentary strata. Subtle scarps on Pleistocene surfaces are visible on high-resolution lidar topography at a number of locations along the aeromagnetic lineament. Collectively, the scarps are coincident with the aeromagnetic lineament and extend a total distance of 18 km. In the field, scarps exhibit 1 to 5 m of northeast-side-up offset. Two trenches were excavated across one of the lidar scarps during the summer of 2004. Both trenches showed evidence for folding, presumably above a buried reverse/oblique fault tip with at least 2 m of vertical offset. One trench also exposed a normal fault, although it was not possible to determine whether slip was caused by glacial or tectonic processes. The radiocarbon age of a folded, buried soil indicates that the earthquake occurred after 12 ka. Thus, the SWIF has produced at least two slip events in the Holocene, one occurring $\sim$3 ka on Whidbey Island and the other $\sim$12 ka on the Washington mainland.
G44A-02 INVITED 16:15h
High resolution (10 km, 3mGal) gravity mapping of the Global continental margins using ERS1 and Geosat satellite altimeter data
Converting sea-surface height variations, derived from satellite altimetry, to free air gravity is not new. What is new is our ability to use existing satellite data to resolve anomalies down to 10 km wavelength and to within 2 to 5 km of the coast globally. Since 1993, GETECH, University of Leeds, has been involved in developing new methods and techniques to recover gravity from satellite altimeter data with the specific aim of improving oil exploration methods in deep-water areas of the continental margins, worldwide. By a combination of new processing methods, we can recover anomalies that are accurately and reliably representing geological structures down to 10 km in wavelength. These improvements have been possible by applying: -advanced seismic repicking methods to recover the onset of ERS-1 radar waveforms; -a range of more accurate and globally consistent wet and dry Troposphere propagation and tidal corrections; -careful data editing and validation based on data visualisation techniques; -micro-levelling to resolve discrepancies between satellite tracks in order to derive a reliable geoid surface; and -the `Geoid to Gravity' conversion method which has significant advantages over existing methods The methodology has allowed us to map to within 2 and 5 km of the coast, globally, and significantly suppress noise that is prevalent and dominates existing solutions at 40 km and shorter wavelengths. At these and shorter wavelengths the geological signal is effectively masked by the noise. The global continental margins study was completed in June 2004 and examples of the resolution improvements for the Gulf of Mexico and S E Asia are shown. Key words: Satellite gravity, high-resolution, continental margins.
G44A-03 INVITED 16:30h
Constructing Full Spectrum Potential-field Anomalies for Enhanced Geodynamical Analysis Through Integration of Surveys From Different Platforms
To complement multi-disciplinary regional geophysical projects such as EarthScope (USA), a significant opportunity exists in potential-fields for improving the characterization of density and magnetic variations inside the Earth and constraining geodynamical parameters such as variation of temperature and strength of the lithosphere. However, these contributions require complete and accurate anomaly spectrum (a few km to the longest possible anomaly wavelength). Recent and continuing upgrade of North American gravity and magnetic anomaly maps, Australian magnetic map, and on-going compilation of the world digital magnetic anomaly map reflect the recognition of defects in present databases and importance of the full spectrum anomaly field in the interpretation of regional geology. In magnetics, problems in adequately collected data stem from variety of sources ranging from defects in secular variation of IGRF models and the consequent need to warp local magnetic surveys to accomplish continental compilations to lack of accurate methods to compare and integrate data sets from disparate platforms (ground, marine, aeroplane, balloon, satellite). Methods have been developed in the last decade (Sabaka et al., 2002, Geophys.J.Int., 151, 32-68; Ravat et al., 2002, Geophysics, 67, 546-554; Ravat et al., 2003, The Leading Edge, 22, 784-785) for processing, comparing, and integrating these disparate data that show that when properly reduced data sets are combined with appropriate methods, it is possible to create full spectrum anomaly fields in many situations. Yet, challenges remain in the intermediate wavelength (150-500 km) magnetic anomaly band over most of the world, especially where original observation parameters are not available. There are similar challenges in the integration of the gravity field. Long wavelengths of the potential-fields are critical for interpreting deep crustal magnetization and even deeper density variations, constraining the temperature field and strength variations of the lithosphere, correctly isolating regional geologic variations from the regional effects such as continent-ocean contrasts, and also isolating core field from lithospheric magnetic fields.
G44A-04 16:45h
3D Joint Inversion of Travel-Time and Potential Field Data Across the Chicxulub Impact Crater
The Chicxulub impact crater lies partly offshore/onshore and is buried beneath the Yucatan Peninsula in Mexico. At 180-200 km in diameter, Chicxulub is one of the three largest known impact craters on Earth, and has been associated with the K-T mass extinction. Chicxulub has been the target of several large-scale geophysical experiments and we now have seismic reflection, seismic refraction, gravity, magnetic and magnetotelluric data across this crater. Several distinct structural and lithological models of Chicxulub have been proposed, and they have been constructed using the geophysical data, a limited amount of borehole data, and observations from other terrestrial craters. Deep crustal rocks have been uplifted to within a few kilometers of the Earth's surface near the crater center, but the precise shape and amount of uplift remains unclear. Mapping this uplift is of importance because it will help us to better understand the kinematics and dynamics of crater formation. In an effort to constrain the shape of this stratigraphic uplift, we are in the process of jointly inverting the 3D tomographic travel-time and potential field data. An initial 3D velocity model has been constructed using the travel-time data and the 3D fast code of Zelt and Barton (1998). Velocity is defined along a regular grid. Velocity models are converted to density models (and vice versa) using a relationship determined from measurements on core taken from a borehole in the central crater basin. We intend to use a joint constrained seismic/gravity 3D velocity inversion in the area of dual coverage and gravity only inversion in the rest of the model. Constraints will include velocity/density bounds and smoothing constraints so that the result fits well with the lab density/velocity measurements and to account for different data resolutions of seismic and gravity data. We will explore a number of inversion approaches, and a range of model parameterizations. Once we are satisfied with our procedure we will attempt to incorporate the magnetic field data into the inversion. Zelt, C. A. and P. J. Barton (1998). Three-dimensional seismic refraction tomography: A comparison of two methods applied to data from the Faroe Basin. Journal of Geophysical Research 103: 7187-7210.
G44A-05 17:00h
Modelling the salar de Uyuni, Bolivia as an equipotential surface of Earth's gravity field
The salar de Uyuni is a massive dry salt lake that lies at the lowest point of an internal-drainage basin in the Bolivian Altiplano. Its topography is remarkable for its extraordinary flatness over almost a full degree of latitude and longitude, despite constant topographic forcing due to faulting and local isostatic rebound. We surveyed a 54 x 45 km region of the salar with kinematic GPS in September, 2002 and found a topographic range of only 80 cm over the surveyed area. Furthermore, the survey revealed distinct surface features with wavelengths between 5 and 40 km. Some of these appear to be aligned with orographic features that intersect the salar, leading us to conjecture that they are the surface expression of high-density massifs that have been buried by low-density basin sediments. Over the oceans, a similar correspondence between basin bathymetry and surface topography is exploited to map the seafloor using sea-surface satellite altimetry measurements, with the sea surface following geoid undulations due to the underwater mass distribution. On the salar, annual flooding creates a shallow seasonal lake whose surface is likewise an equipotential shaped by the distribution of underlying mass. The dissolution and redeposition of salt by the lake waters appears to push the system toward an equilibrium of constant water depth so that the salt surface itself closely approximates the local equipotential surface. To test our hypothesis about the origin of the surface features on the salar, we compare our GPS survey elevations with the equipotential surface estimated from a combined analysis of local gravity measurements and the EGM96 global geopotential model. 50% of the variance of the GPS elevations can be explained by equipotential surface undulations from the EGM96 model alone. An additional 40% is explained by the shorter-wavelength equipotential surface derived from local gravity. The elevation residual is remarkably consistent with our independent estimate (from remote-sensing analysis of water depths) of the departure of the surface from an equipotential, suggesting that it is the signature of unmodelled secondary surface processes. We believe that the salar de Uyuni is one of the few terrestrial locations where potential fields are able to play a major role in shaping local ($<$ 100 km) topography, overcoming the typical dominance of tectonics and erosion in geomorphology on this scale.
G44A-06 17:15h
On the Inversion for Mass (Re)Distribution from Global (Time-Variable) Gravity Field
The well-known non-uniqueness of the gravitational inverse problem states that the external gravity field, even if completely and exactly known, cannot uniquely determine the density distribution of the body that produces the gravity field. In this paper we provide conceptual insight by examining the problem in terms of spherical harmonic expansion of the global gravity field. By comparing the multipoles and the moments of the density function, we show that in 3-D the degree of knowledge deficiency in trying to inversely recover the density distribution from external gravity field is (n+1)(n+2)/2 - (2n+1) = n(n-1)/2 for each harmonic degree n. On the other hand, on a 2-D spherical shell we show via a simple relationship that the inverse solution of the surface density distribution is unique. The latter applies quite readily in the inversion of time-variable gravity signals (such as those observed by the GRACE space mission) where the sources largely come from the Earth's surface over a wide range of timescales.
G44A-07 17:30h
Using high-resolution aeromagnetic survey to map tectonic elements of plate boundaries: An example from the Dead Sea Fault
The Dead Sea Fault (DSF) is a transform plate boundary between the African and the Arabian plates. The 200-km-long DSF segment between the Gulf of Aqaba/Elat and the Dead Sea, which has the morphology of a rift valley, shows little seismic activity, and its surface trace is only intermittently visible. High-resolution magnetic data were collected in October 2003 aboard a Jordanian military helicopter flying at an altitude of 100 m over the southern 120-km-long section of this fault segment. The survey was part of a US-AID Middle Eastern Regional Cooperation project between Jordanian, Israeli, Palestinian, and American scientists. Data were collected along rift-perpendicular lines spaced 300 m apart, requiring frequent crossings between Israeli and Jordanian air spaces. The data were gridded at 75 m interval following resolution tests, reduced to pole, and incorporated into a GIS together with elevation, geology, and gravity maps to facilitate interpretation. The main findings of the magnetic survey are the absence of magnetic anomalies crossing the rift valley, and the presence of a rift-parallel regional lineament corresponding to the active trace of the DSF. The lineament extends NNE as an almost continuous trace from Elat, Israel, to the eastern side of the valley 5 km north of Rahmeh. Jordan. Another fault trace located 2-3 km to the west may overlap and continue NNE through Gebel A-Risha, and into the central Arava/Araba valley, where it is visible on the surface. Alternatively, the two traces may be connected. If an offset between the two traces exists, it may be small enough to allow an earthquake rupture to propagate across the offset, and generate an earthquake with a moment magnitude of up to 7.5. Traces of buried faults in the central Arava/Araba valley that were previously active in the DSF system, are visible as abrupt terminations of an area of short wavelength magnetic anomalies. These anomalies probably represent shallow subsurface magmatic intrusions. The closest exposed intrusion is dated at 20.7 Ma, shortly before the development of the DSF. Other anomalies can be traced at the edges of our survey area and are likely related to Precambrian outcrops along the rift shoulders. Comparison of the magnetic and the sparser land-gravity data shows the same general azimuth of the magnetic lineament and of the segmented fault system as derived from the gravity and a surprisingly good coincidence between local gravity and magnetic anomalies over the Timna pull-apart basin, owing perhaps to the sensitivity of the high-resolution magnetic data to the thickness of the sedimentary cover.
G44A-08 17:45h
An Irregular-gridded Stable Potential-field Downward Continuation Method
Potential-fields downward continuation can increase the resolution, while it is an inherent ill-posed inverse problem. We advance a fast algorithm to solve the interpolation coefficients of arbitrary-spaced four variable cubic B-spline. The downward continuation, both 2D and 3D, is accomplished by solving integral equations using B-spline bases in space domain. In contrast to FFT method, our method can be irregular spacing, and the number of knots need not to be a power of 2. Through comparison with FFT method using synthetic examples, including noise-contaminated data continuation, it is found that our method is more accurate and more stable. Real data applications of B-spline method downward continuation provide very useful information for further interpretation.