HR: 13:40h
AN: V53F-01 INVITED [Abstracts]
TI: Recycling volatiles and attaining a geochemical and fluid dynamically consistent model of mantle
convection
AU: * Ballentine, C J
EM: chris.ballentine@manchester.ac.uk
AF: University of Manchester, School of Earth, Atmosphere and Environmental Science, Manchester, M13 9PL
United Kingdom
AU: Van Keken, P E
EM: keken@umich.edu
AF: University of Michigan, Dept. Geological Science, Ann Arbor, MI MI 48109
United States
AU: Holland, G
EM: greg.holland@manchester.ac.uk
AF: University of Manchester, School of Earth, Atmosphere and Environmental Science, Manchester, M13 9PL
United Kingdom
AU: Hauri, E H
EM: hauri@dtm.ciw.edu
AF: Carnegie Institute of Washington, Dept. Terrestrial Magnetism, Washington, DC DC 20015
United States
AU: Brandenburg, J
EM: jpbrande@umich.edu
AF: University of Michigan, Dept. Geological Science, Ann Arbor, MI MI 48109
United States
AB:
Numerical simulations of mantle convection provide a critical test of models advocated to account for the geochemical
observables, and in particular the noble gas isotope variation between Ocean Island Basalts (OIB) and Mid Ocean Ridge Basalts
(MORB). For the first time, an unambiguous data set describing the convecting mantle (MORB-source) noble gas isotopic
composition and relative elemental abundance of He/Ne/Ar/Kr [1] and now Xe [2] has been derived from magmatic CO2 natural
gases. The simplest explanation for the non-radiogenic 36Ar/84Kr/130Xe abundance pattern is recycling of unfractionated
seawater into the convecting mantle. By mass balance, this has negligible effect on the He/Ne system. We have tested this
conceptual model by modifying numerical models of mantle convection [3] to input noble gases with seawater relative
abundances and isotopic composition into a mantle initially dominated by residual Solar gases and noble gases produced by the
decay of U, Th and K over time. The flux of seawater into the model mantle has been matched to the in-situ production of
40Ar to give an average convecting mantle 40Ar/36Ar~46,000 [2], sampled at ridge analogues, at the end of the model run
simulating 4.5Ga. The consequence of all models has been to produce a region of the mantle at the core-mantle boundary
significantly more enriched in 36Ar, 84Kr and 130Xe relative to radiogenic noble gas isotopes. Model plumes upwelling from
this region preserve this signature and are qualitatively consistent with the observed OIB 40Ar/36Ar and 129Xe/130Xe maxima
of ~10,000 and <7.0 respectively. Initial models that most closely match the observed values require `strong' (more
viscous) recycled slabs, introduced into the model by considering temperature controlled viscosity. These models do not yet
account for the plume source high 3He/4He and we are exploring other mechanisms in the numerical models that may introduce or
preserve a more primitive volatile signature, such as a deep (core or D'') 3He source or preservation of high 3He/4He during
the recycling process [4] - and still preserve the atmosphere-radiogenic isotope difference observed between OIB and MORB.
[1] Ballentine, C. J., et.al. Nature 433, 33-38, 2005. [2] Holland. G and C.J. Ballentine, AGU abstract, This meeting. [3]
Van Keken, P.E., et. al., The Mantle and Core, 471-492, Elsevier, 2003. [4] Class, C. & Goldstein, S. L. Nature 436,
1107-1112, 2005
DE: 1009 Geochemical modeling (3610, 8410)
DE: 1025 Composition of the mantle
DE: 1038 Mantle processes (3621)
DE: 1040 Radiogenic isotope geochemistry
SC: Volcanology, Geochemistry, Petrology [V]
MN: Fall Meeting 2005