Comparison of Marine Gravity from Shipboard and High-density Satellite Altimetry along the Mid-atlantic Ridge, 30.5ø-35.5øs

We compare new marine gravity fields derived from satellite altimetry with shipboard measurements over a region of more than 120,000 square kilometers in the central South Atlantic. Newly declassified satellite data were employed to construct free-air anomaly maps on 0.05 degree grids [Sandwell and Smith, 1992; Marks et al., 1993]. An extensive gravity and bathymetry dataset from four cruises along the Mid-Atlantic Ridge from 30.5-35.5øS provides a benchmark for testing the two-dimensional resolution and accuracy of the satellite measurements where their crosstrack spacing is near their widest. The satellite gravity signal is coherent with bathymetry in this region down to wavelengths of 26 km (T2=0.5), compared to 12.5 km for shipboard gravity. Residuals between the shipboard and satellite datasets have a roughly normal distribution. The standard eviation of satellite gravity with respect to shipboard measurements is nearly 7 mGal in a region of 140 mGal total variation, whereas the internal standard deviation at crossovers for GPS-navigated shipboard data is 1.8 mGal. The differences between shipboard and satellite data are too large to use satellite gravity to determine crustal thickness variations within a typical ridge segment. Introduction Three-dimensional analysis of gravity and bathymetry provides a unique look at the density structure of crust and upper mantle at mid-ocean ridges [Kuo and Forsyth, 1988; Lin et al., 1990; Blackman and Forsyth, 1991; Morris and Detrick, 1991]. Maps of the marine gravity field [Sandwell, 1992; Sandwell and Smith, 1992; McAdoo and Marks, 1992; Marks et al., 1993; Smith et al., 1993] based on Seasat, ERS-1, and recently declassified Geosat Geodetic Mission (GM) altimetry in the southern hemisphere from 30øS-72øS, together with improvements in the density and quality of the marine bathymetric database, will enhance the study of gravity variations on the global ridge system. The purpose of this note is to assess the current state of the art in marine gravimetry by a two-dimensional (2-D) comparison of the free-air gravity anomaly (FAA) field derived from satellite altimetry with that obtained from shipboard measurements, in an area extending nearly five degrees in latitude and longitude. Satellite Altimetry and Shipboard Gravity Datasets Sandwell (Eos, Jan. 19, 1993, p.35; updated Mar. 18, 1993) has generated a global marine gravity field using 20 pixels per degree in longitude, with the same pixel dimensions in latitude, on a Mercator projection. The spacing of this grid is about 4.5 km over the Mid-Atlantic Ridge (MAR) from 30.5ø-36.5øS and 12ø-17øW. The vertical gravity gradient is calculated using Laplace's equation for gravitational potential Copyright 1993 by the American Geophysical Union. Paper Number 93GL01487 0094-8534/93/93 GL-01487503.00 from gridded maps of the vertical deflection (i.e. sea surface slope) in the east and north directions, as described in Sandwell [1992]. The gravity anomaly, relative to a spherical harmonic model, is then generated via Fourier analysis, using a flat-earth approximation. These steps are aided by an efficient POCS-like interpolation algorithm [Menke, 1991 ] for the along-track vertical deflections. Marks et al. [1993] also produced a gravity field for the southern oceans, using Geosat data only, spaced 0.05 ø in longitude, 0.04 ø in latitude, registered on grid lines. The accuracy of this grid is somewhat improved over their earlier version by the optimal filtering techniques of Smith et. al. [1993], which cannot as yet be applied to data from ERS-1. This grid is now available from NOAA through the National Geophysical Data Center (NGDC Data Announcement 93MGG-01, Global Relief Data on CD-ROM, 1993). Figure 1 shows the satellite tracks now available, as well as the ship tracks in the area used for the comparison. The