ALBERT

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution June 1993

Description: This dissertation studies several aspects of the formation of the Earth's oceanic mantle and crust, using a variety of geologic techniques, principally major elements, radiogenic isotopes and trace elements, but including petrography, mineral chemistry, x-ray diffraction, seafloor geomorphology, and analysis of the tectonics of fracture zones. The first chapter is an introduction to the problems to be addressed in this work. The second chapter examines the composition of basalts erupted near the Atlantis II Fracture Zone on the Southwest Indian Ridge. Trends in major element compositions of those basalts can be related directly to the nearby presence of the fracture zone. The effects of mantle composition and crustal level lateral transport of magma in the rift system can be ruled out by the analysis of isotopes and the geomorphology of the fracture zone floor. This is the best demonstration to date of a transform fault effect on basalt compositions. In trying to quantify putative transform fault effects documented at other fracture zones, no systematic correlation of transform offset age with mantle temperature change can be found, suggesting that mantle composition and lateral transport phenomena play a larger than expected role in the evolution of those areas. The third chapter relates to oceanic mantle rocks as they are altered at or near the Earth's surface. The major elements which make up abyssal peridotites are extensively redistributed by the alteration they have undergone. Mg is shown to be extracted from the peridotites, and a variety of trace elements added. This elemental redistribution is taken as evidence for extensive Mg transport by circulating waters. Since the solubility of Mg-bearing minerals in hydrothermal solutions is quite limited, much lower temperatures and much higher water /rock ratios are required to explain the major element compositions of the peridotites than had previously been assumed. The behavior of the Nd, Sr and Os isotopic systems during seafloor alteration was also studied. The isotope systematics of these rocks strongly support the hypothesis of high water /rock ratios in the formation of serpentinized abyssal peridotites. Nonetheless, Nd and Sr reside in a phase which is resistant to alteration (clinopyroxene) and the concentration of Os is high relative to that of seawater, so that it too appears resistant to alteration. Primary mantle isotopic signatures may be obtained from abyssal peridotites by careful analysis, even of extremely weathered rocks. Radiogenic strontium in excess of what could be introduced by seawater contamination or in situ radiogenic growth in a reasonable period of time was also found. These observations confirm earlier work which had been discredited for many years. The only plausible mechanism for the formation of this "orphan" S7Sr is that it is introduced as part of a sedimentary component which infiltrates the rock during metamorphism and/ or weathering. The 87Sr may be contained by or sorbed onto extremely fine clay particulates, or colloidal suspensions, as opposed to the dissolved ionic Sr which is normally thought of as characterizing the Sr isotopic composition of seawater. The high water/rock ratios required by the bulk isotopic analysis, as well as the pervasive elemental redistribution arguing for extensive near-surface weathering at high water /rock ra.tios strongly support this hypothesis. Given pervasive percolation of water throughout the samples, sufficient radiogenic, sediment-derived strontium may be drawn deep into the crust in the course of its weathering to cause such high B7SrfB6Sr ratios. The fourth chapter deals exclusively with primary mantle isotopic information from abyssal peridotites. This is the first study which has attempted to relate the Os isotopic system in the oceanic mantle to other isotopic systems and to trace elements. It is possible, with some extreme assumptions, to model the range of Os isotopes in the oceanic mantle alone in a standard model of formation of the depleted mantle by extraction of the crust. The additional constraints provided by the study of Nd isotopes in depleted mantle rocks from the oceans show that partial melt extraction and the formation of a depleted reservoir alone are not sufficient to account for the range of both N d and Os isotopes in the Earth's mantle. Possible mechanisms for the decoupling of the Os and Nd isotopic systems include elemental fractionation via the porous flow of basalt through the mantle, mantle metasomatism, recycling of a subducted component in the mantle and core formation. The core extraction model is pursued in some detail. Such core extraction models can account for the distributions and isotopic compositions of compatible and incompatible trace elements in the Earth's mantle, but they are highly non-unique, and thus difficult to test.

Description: This thesis was supported by an NSF Graduate Fellowship, and by a grant from the WHOI Ocean Ventures Fund.

Description: Author Posting. © American Geophysical Union, 2004. This article is posted here by permission of American Geophysical Union for personal use, not for redistribution. The definitive version was published in Geochemistry Geophysics Geosystems 5 (2004): Q08002, doi:10.1029/2004GC000712.

Description: Detailed hydrothermal surveys over ridges with spreading rates of 50–150 mm/yr have found a linear relation between spreading rate and the spatial frequency of hydrothermal venting, but the validity of this relation at slow and ultraslow ridges is unproved. Here we compare hydrothermal plume surveys along three sections of the Gakkel Ridge (Arctic Ocean) and the Southwest Indian Ridge (SWIR) to determine if hydrothermal activity is similarly distributed among these ultraslow ridge sections and if these distributions follow the hypothesized linear trend derived from surveys along fast ridges. Along the Gakkel Ridge, most apparent vent sites occur on volcanic highs, and the extraordinarily weak vertical density gradient of the deep Arctic permits plumes to rise above the axial bathymetry. Individual plumes can thus be extensively dispersed along axis, to distances 〉200 km, and ∼75% of the total axial length surveyed is overlain by plumes. Detailed mapping of these plumes points to only 9–10 active sites in 850 km, however, yielding a site frequency F s , sites/100 km of ridge length, of 1.1–1.2. Plumes detected along the SWIR are considerably less extensive for two reasons: an apparent paucity of active vent fields on volcanic highs and a normal deep-ocean density gradient that prevents extended plume rise. Along a western SWIR section (10°–23°E) we identify 3–8 sites, so F s = 0.3–0.8; along a previously surveyed 440 km section of the eastern SWIR (58°–66°E), 6 sites yield F s = 1.3. Plotting spreading rate (us) versus F s, the ultraslow ridges and eight other ridge sections, spanning the global range of spreading rate, establish a robust linear trend (F s = 0.98 + 0.015us), implying that the long-term heat supply is the first-order control on the global distribution of hydrothermal activity. Normalizing F s to the delivery rate of basaltic magma suggests that ultraslow ridges are several times more efficient than faster-spreading ridges in supporting active vent fields. This increased efficiency could derive from some combination of three-dimensional magma focusing at volcanic centers, deep mining of heat from gabbroic intrusions and direct cooling of the upper mantle, and nonmagmatic heat supplied by exothermic serpentinization.

Description: This research was partially supported the NOAA VENTS Program. P.J.M. and H.J.B.D. gratefully acknowledge NSF grant OPP 9911795 for support of the AMORE Expedition; P.J.M. and E.T.B. acknowledge NSF grant OPP 0107767 and the VENTS Program for development and construction of MAPRs for use in ice-covered seas. H.J.B.D. acknowledges NSF grant OCE-9907630 for support of SWIR studies. J.E.S. was supported by Deutsche Forschungsgemeinschaft grant SN15/2.