A three‐dimensional gravity analysis of the East Pacific Rise from 18° to 21°30′S

Abstract

Multibeam bathymetry and gravity coverage of the East Pacific Rise (EPR) between 18 ø and 21 ø30'S is used to investigate the relation between melt supply and tectonic segmentation at ultrafast spreading rates. The long-wavelength features in the residual anomaly show a good correlation with those in the bathymetry. The highest residual anomaly values occur over the broad discordant zone of the 20ø40'S overlapping spreading center (OSC), for seafloor ages of 0 Ma to at least 1.5 Ma. We interpret the deepening of the bathymetry and the increase of the residual anomaly toward that discordant zone as due to a decrease of 500 + 200 m of the crustal thickness. Hence the 20ø40'S OSC has been associated with a reduced magmatic budget for at least the past 1.5 m.y. and represents a persistent segmentation of the EPR. This is consistent with models in which mantle upwelling, even at the fastest spreading centers, is enhanced between large discontinuities rather than evenly distributed along axis. However, this decrease of the crustal thickness toward an axial discontinuity is several times smaller than that typically documented for slow spreading ridges, which suggests that mantle upwelling is less focused at fast spreading ridges, or that along-axis transport of crustal material is more efficient, or both. Across the study area, the residual anomaly decreases toward the NW by 15-20 mGal. This regional gradient can be modeled with lateral temperature variations in the upper mantle of up to 60øC, increasing toward the NW. This interpretation is consistent with the numerous seamounts present to the NW and the robust magmatic budget of the ridge between 17 ø and 18øS, and it could also explain why the ridge segments defined by the smaller OSCs between 18 ø and 19øS propagate very rapidly away from the robust area. Similar patterns of ridge propagation away from the shallowest section of a ridge have been documented near the Galapagos and Easter Island hot spots. Hence these shorter ridge segments may not be associated with significant individual melt sources. Rather, they may represent a superficial segmentation due to the interaction between the EPR and a mantle heterogeneity located between 17 ø and 18øS. Macdonald et al., 1984; Schouten et al., 1985; Crane, 1985; Macdonald et al., 1988a]. An implication of focused mantle upwelling models is that the magma supply should decrease toward large discontinuities. Indeed, at slow spreading ridges, seismic refraction data have shown that the crust accreted near •Now at Lamont-Doherty Earth Paper number 95JB00243. 0148-0227/95/95JB-00243505 .00 transform zones can have as little as half the typical oceanic crustal thickness [e.g., White et al., 1984]. At fast spreading ridges, detailed analyses of multibeam bathymetry data show that the deepening of the axial high toward large overlapping spreading centers (OSCs) and transform faults correlates with a narrowing of its cross section and, often, with the disappearance of the seismic reflector interpreted as the top of an axial magma...

The influence of magma supply and eruptive processes on axial morphology, crustal construction and magma chambers

“…This is consistent with the small along-axis variations in crustal thickness inferred from gravity data in this area [Cormier and Macdonald, 1995;Magde et al , 1995]. Diapiric feeding of fast spreading ridge axes has recently been suggested by Nicolas et al [ 1996] and Wang et al [1996].…”

“…The ridge topography is supported by a low density volume underlying the ridge crest, which is partitioned between the crust and the mantle [Wang and Cochran, 1993;Magde et al, 1995]. The along-axis variation in gravity (range = 20 mGal) and part of the axial depth changes (range = 400 m) can be explained by a decrease in crustal thickness of -500 m between 18°S and 20°S [Cormier and Macdonald, 1995]. The variation in cross-sectional area (range= 6 km2) suggests a change in volume of the hot, low density region in the lower crust accompanied by a variation in the density anomaly in a narrow, deep region in the mantle [Magde et al , 1995].…”

Ridge segmentation, tectonic evolution and rheology of slow-spreading oceanic crust

“…The "bull's eye" gravity lows, however, are of lesser amplitude at segments of fast-spreading ridges such as the East Pacific Rise (EPR) [Madsen et al, 1990;Wang and Cochran, 1993;Cormier et al, 1995] Morgan, 1990; suggest a transition from three-dimensional plumelike mantle upwelling to two-dimensional sheetlike upwelling with increasing spreading rate which is consistent with the observed spreading rate dependence of ridge axis gravity structure . The range of MBA amplitudes decreases with increasing spreading rate (Figure 2).…”

“…The existence of distinctive lows in mantle Bouguer gravity anomalies (MBA) at the Mid-Atlantic Ridge (MAR) (Figure 1) indicate that the crust is thicker and the upper mantle is hotter at segment centers than at distal ends [Kuo and Forsyth, 1988;Blackman and Forsyth, 1991;Morris and Detrick, 1991;. These residual gravity lows suggest that magmatic accretion occurs in the form of discrete, buoyancydriven upwelling centers [Kuo and Forsyth, 1988;, thus supporting earlier hypotheses on the existence of melt or buoyant mantle diapirs in the underlying asthenosphere [e.g., Rabinowicz et al, 1984;.The "bull's eye" gravity lows, however, are of lesser amplitude at segments of fast-spreading ridges such as the East Pacific Rise (EPR) [Madsen et al, 1990;Wang and Cochran, 1993;Cormier et al, 1995] Morgan, 1990; suggest a transition from three-dimensional plumelike mantle upwelling to two-dimensional sheetlike upwelling with increasing spreading rate which is consistent with the observed spreading rate dependence of ridge axis gravity structure . The range of MBA amplitudes decreases with increasing spreading rate (Figure 2).…”

Constraints on a buoyant model for the formation of the axial topographic high on the East Pacific Rise

“…These rates are somewhat lower than estimated by Cochran [ 1986], probably because his profiles extend much farther from the axis than our survey and the approximations of the cooling half-space thermal model tend to overestimate the cooling rate near the ridge axis [Cormier et al, 1995] …”