We have used a unique data set collected during the 1985 East Pacific Rise multichannel seismic experiment to reevaluate the constraints that gravity data can place on the crustal structure of the East Pacific Rise (EPR). The close spacing of track lines within the two main survey areas (8¿45'N--9¿55'N and 12¿20'--13¿30'N) allowed us to perform three-dimensional analyses of high-quality gravity data (+/- 1 mGal uncertainity) obtained with the Bell Aerospace BGM-3 gravity meter. Our gravity modeling was enhanced by the availiability of high-resolution Sea Beam bathymetric data and by independent structural constraints provided by seismic reflection and refraction data. To model the crustal and upper mantle density structure at the ridge axis, we first calculated the gravity anomalies due to the density contrasts at the water/crust and crust/mantle boundaries and the changes in density caused by the cooling of the lithosphere with age and substracted these predictable signals from the observed free-air anomaly.

The residual anomalies were then used to place constraints on the magnitude and distribution of anomalous mass at the EPR. Our results show that over 90% of the power in the observed free-air anomaly can be modeled by these predictable components of the gravity signal. The amplitude and wavelength of the small residual anomaly can be modeled by a broad region (~20 km wide), centered on the rise axis, of slightly lower-than-normal crustual and/or upper mantle densities (-0.03 Mg m-3). This region of anomalous mass and density variations in the underlying mantle provide the principle isostatic support for the axial topographic high. The anomalies are consistent with, but do not require, the presence of a low-density, crustal magma chamber at the rise axis. If a largely molten magma chamber exists then the gravity data require that it be a narrow, volumetrically small body.

There are first order differences in the distribution of anomalous mass between the two survey areas with the 9¿N region generally being characterized by lower densities in the axial region. Along-axis variations in anomalous mass within each of the survey areas correlate with first-order changesa in the depth and cross-sectional shape of the EPR axial high. We attribute these differences in axial morphology and the amplitude of the associated gravity anomalies to along-strike change in the size of the crustal magma chamber, the width of the surrounding zone of cooling crustal rocks, and densities in the uppermost mantle beneath the rise axis. ¿ American Geophyical Union 1990