Engineering topics
Publications and source records attributed to Sandwell, David T..
Publications related to geophysical applications of Seasat and Geosat altimetry are reviewed for the period 1987-1990. Problems discussed include geoid and gravity errors, regional geoid heights and gravity anomalies, local gravity field/flexure, plate tectonics, and gridded geoid heights/gravity anomalies.
Satellite altimeter data from the first 44 repeat cycles (2 years) of the Geosat Exact Repeat Mission (EWRM) were averaged to improve accuracy, resolution and coverage of the marine gravity field. Individual 17-day repeat cycles were first edited and differentiated, resulting in the along-track vertical deflection (i.e., gravity disturbance). To increase the signal-to-noise ratio, 44 of these cycles were then averaged to form a single highly accurate vertical deflection profile. The largest contribution to the vertical deflection error is short-wavelength altimeter noise and longer-wavelength oceanographic variability; the combined noise level is typically 6 microrad. Both types of noise are reduced by averaging many repeat cycles. Over most ocean areas the uncertainty of the average profile is less than 1 microrad which corresponds to 1 mgal of along-track gravity disturbance. However, in areas of seasonal ice coverage, its uncertainty can exceed 5 microrad. To assess the resolution of individual and average Geosat gravity profiles, the cross-spectral analysis technique was applied to repeat profiles. Individual Geosat repeat cycles are coherent (greater than 0.5) for wavelengths greater than about 30 km and become increasingly incoherent at shorter wavelengths.
A simple derivative filter technique for extracting mesoscale variability from satellite altimeter data having large-amplitude, long-wavelength errors, with no data adjustments necessary, is presented. The method is applied to the first year of Geosat Exact Repeat Mission data. The main advantage over previous filter methods is that the derivative filter is only two points long, so that it produces no edge effects. Short profiles can thus be included in the analysis. The derivative filter also transforms a dynamic height profile into a dynamic slope profile. The dynamic slope of the sea surface is a parameter that can be measured independently. Moreover, if the ocean current is in geostrophic balance, then the dynamic gradient of the sea surface is proportional to the current velocity.
A total of 44 Geosat profiles over ridges with spreading rates ranging from 14 to 155 mm/yr were analyzed. In agreement with previous studies, it is found that slow spreading ridges usually have high amplitude gravity troughs, while fast spreading ridges are characterized by low-amplitude ridge axis highs. Unexpectedly, it is found that the transition from axial trough to axial high occurs abruptly at a spreading rate of 60-70 mm/yr. Ridge axis gravity signatures are highly variable for rates less than 65 mm/yr and very uniform at higher rates. The transition of the gravity signature appears to get more abrupt than the transition of the topographic signature, suggesting an abrupt change in the style of isostatic compensation with spreading rate. Published models of ridge axis dynamics do not explain this sharp transition.
A model of the tectonic evolution of the eastern Indian Ocean from the Late Cretaceous to the present is constructed. Reinterpretations of seafloor magnetic anomalies are used to place constraints on spreading rates and the timing of major reorganizations. Vertical deflection profiles from 22 repeat cycles of Geosat altimeter data are used to determine the tectonic fabric associated with fracture zones. The Geosat data place constraints on paleospreading directions. Consideration is given to the implications of the model for the relative motions between India, Australia, and Antarctica.
The source of cross-grain lineations in marine gravity field observed in central Pacific was investigated by comparing multiple collinear gravity profiles from Geosat data with coincident bathymetry profiles, in the Fourier transform domain. Bathymetric data were collected by multibeam sonar systems operating from two research vessels, one in June-August 1985, the other in February and March 1987. The results of this analysis indicate that the lineations are superficial features that appear to result from a combination of subsurface and surface loads supported by a thin (2 km to 5 km) lithosphere.
Gridded geoid height data (Marsh et al.l, 1986) and gridded bathymetry data (Van Wykhouse, 1973) are used to estimate the average compensation depths of 53 oceanic swells and plateaus. The relationship between geoid height and topography is examined using Airy and thermal compensation models. It is shown that geoid height is linearly related to topography between wavelengths of 400 and 4000 m as predicted by isostatic compensation models. The geoid/topography ratio is dependent on the average depth of compensation. The intermediate geoid/topography ratios of most thermal swells are interpreted as a linear combination of the decaying thermal swell signature and that of the persisting Airy-compensated volcanic edifice.
Geosat altimeter data, collected from an orbit with a ground rack that repeated every 17 days and overlayed one of the 17-day Seasat ground tracks, were used to map the gravity field of the Southern Ocean and the continental margin of Antarctica. The combination of ascending an descending profiles produced a typical Geosat ground track spacing of 70 km at the equator, with the best coverage occurring between the latitudes of 60 and 72 deg in both the Northern and Southern hemispheres. The new data reveal many previously uncharted seamounts and fracture zones in the extreme Southern Ocean areas adjacent to Antarctica, showing the detailed gravity signatures of the passive and active continental margins of Antarctica. Seven large age-offset fracture zones apparent in the Geosat data record the early breakup of Gondwana.
A new measurement techique based on a model of a Gaussian seamount loading a thin elastic lithosphere was developed to analyze seamounts that, until then, were not surveyed or seamounts with poor bathymetric coverage. The model predicts that the seamount diameter is equal to the peak-to-trough distance along the vertical deflection profile and that the flexural diameter of a seamount is related to the age of the lithosphere when the seamount formed. This model also suggests that these two measurements are relatively insensitive to the cross-track location from the seamount. These model predictions were confirmed using Seasat altimeter profiles crossing 14 surveyed seamounts in the Pacific. The analysis of the seamount distribution indicated considerable variations in population density and type across the oceans. Most notable among them are the absence of seamounts in the Atlantic, variations in population density across large fracture zones in the Pacific, and the prevalence of small signatures in the Indian Ocean.
An algorithm is presented for determining the minimum curvature surface passing through a set of nonuniformly spaced data points. The curve is generated as a linear combination of Green functions for the biharmonic operator at each data point, with the amplitudes of the functions adjusted so that the interpolating surfaces passes through each point. The function passing through the points can be regarded as a spline to which point forces are applied, defining the minimum curvature between the points. The technique was used to combine the along track slopes of the GEOS-3 and Seasat altimeter data into a consistent geoid height map of the Caribbean area, covering 0.5 million data points in the process. Sample images are provided and new topographic features that are revealed are discussed.