HR: 11:20h
AN: V22A-05    [Abstracts]
TI: Distribution of Water in Earth's Mantle - Implications from Samoan Submarine Lavas
AU: * Workman, R K
EM: rworkman@whoi.edu
AF: Woods Hole Oceanographic Institution, 266 Woods Hole Rd., Woods Hole, MA 02540 United States
AU: Hauri, E
EM: hauri@dtm.ciw.edu
AF: Carnegie Institute of Washington, 5241 Broad Branch Rd. NW, Washington, DC 20015 United States
AU: Hart, S R
EM: shart@whoi.edu
AF: Woods Hole Oceanographic Institution, 266 Woods Hole Rd., Woods Hole, MA 02540 United States
AU: Wang, J
EM: jwang@dtm.ciw.edu
AF: Carnegie Institute of Washington, 5241 Broad Branch Rd. NW, Washington, DC 20015 United States
AU: Blusztajn, J
EM: jblusztajn
AF: Woods Hole Oceanographic Institution, 266 Woods Hole Rd., Woods Hole, MA 02540 United States
AB: We report volatile and trace element data for submarine, basaltic glasses from the three youngest Samoan volcanoes, Ta'u, Malumalu and Vailulu'u - some of which define (by Sr-Nd-Pb isotope compositions) the enriched mantle endmember, EM2 (Zindler and Hart, 1986). Shallow degassing has affected CO₂ in all samples, and H₂O only in the most shallowly erupted samples from Vailulu'u. Absolute water contents are high for Samoa (0.63 - 1.50 wt%), especially compared to MORB glasses which commonly contain 0.2 - 0.5% H₂O. During mantle melting, the compatibility of water is similar to Ce and La (Michael, 1995; Danyushevsky et al., 2000; Hauri et al., in revision). As such, ratios of H₂O/Ce and/or H₂O/La have been used as direct indicators of mantle source compositions. In Samoa, H₂O/Ce (58 - 157) and H₂O/La (120 - 350) are all lower than in MORBs and correlate inversely with ⁸⁷Sr/⁸⁶Sr compositions (0.7045 - 0.7089); this relationship extends the observation by Dixon et al. (2002) that enriched mantle sources have relative depletion of water. Ultimately, it seems that the greater the "enrichment" of the mantle (in terms of heavy isotopes and trace elements), the greater the apparent dehydration. It has been known for decades that the diffusion rates of most lithophile trace elements are so slow that mantle heterogeneities will not equilibrate for these elements over more than 10 meters during the whole age of the Earth (Hofmann and Hart, 1978). On the other hand, experimental data for the diffusion of hydrogen in upper mantle minerals indicate that diffusive mobility of water is extremely rapid. For example, using the information for diffusion of hydrogen in olivine from Mackwell and Kohlstedt (1990), an entire 20 km thick EM2 slab, with initial water content of ~400 ppm, will dehydrate to <110 ppm water in 1 billion years, well within the estimated age of EM2 (~2.5 Ga). Diffusion rates of water in diopside are about 5 times faster than olivine at 1600°C (Woods et al., 2000), so will help to speed the slab's equilibration times. Also, the effects of enhanced diffusion through grain boundaries will serve to aid equilibration. This example calculation shows that chemical equilibration of water in the mantle is a nearly unavoidable consequence of recycling, unless slab temperatures stay very cold (much less than 1600°C), recycling times are very rapid (< 50 million years), or heterogeneities are very large. As a result, the enhanced solid-state mobility of hydrogen can potentially shift H₂O/La ratios in the absence of any melting or mantle-fluid interactions. Therefore, differences in the H₂O/La ratios of Samoan mantle components are due to differences in REE enrichment at fairly constant H₂O. We expect that subducted slabs will experience a two-stage dehydration history, first within subduction zones and then in the ambient mantle during long-term convective mixing. If the diffusion rates of hydrogen in high-pressure mantle minerals are similar to those at low-pressure, water may be one of the only elements whose abundance is nearly constant over great distances in the mantle, assuming similar source lithologies. This "diffusive dehydration" model has important implications for the interpretation of mantle H₂O/La and H₂O/Ce ratios.
DE: 1038 Mantle processes (3621)
DE: 8124 Earth's interior: composition and state (1212, 7207, 7208, 8105)
DE: 8137 Hotspots, large igneous provinces, and flood basalt volcanism
DE: 8410 Geochemical modeling (1009, 3610)
DE: 8430 Volcanic gases
SC: Volcanology, Geochemistry, Petrology [V]
MN: Fall Meeting 2005