Geochemistry, Geophysics, Geosystems

Cover image for Vol. 18 Issue 4

Impact Factor: 2.993

ISI Journal Citation Reports © Ranking: 2015: 17/81 (Geochemistry & Geophysics)

Online ISSN: 1525-2027

New model to estimate water in the Earth's mantle

The Earth's mantle could be holding as much water as the ocean and atmosphere combined but in the form of defects in dry minerals, such as Olivine, that make up a large fraction of the mantle. Water in mantle minerals weakens chemical bonds, affecting physical and chemical properties of the mantle, particularly viscosity and electrical conductivity. Scientists continue debating not just the amount of water but also its distribution in the mantle. Seismic velocity profiles, typically used to map the water content of the mantle, are, in fact, not sensitive to the amount of water. An increase of 1% by weight of water in mantle minerals, i.e., from 'dry' to 'very wet' conditions, reduces seismic velocities by at most 1%, which is often below the detection limit. On the other hand, an increase in water content by only 0.01%, from 'dry' to 'damp' conditions, increases electrical conductivity in mantle minerals by an order of magnitude or more, making electrical conductivity a more sensitive tool to estimate the water content of mantle minerals. Laboratory-derived models used to estimate electrical conductivity through the mantle not only fail to replicate geophysical and petrological observations but also have large disparities between models used by different laboratories. As a result, although electrical conductivity models are sensitive, the scientific community is skeptical of them. Jones et al. (2012) calibrated laboratory-derived models of electrical conductivity of mantle minerals with well-constrained petrological observations of water content from two sites on the cratonic lithosphere in South Africa. But unlike previous models, the authors use a statistical approach, called Monte Carlo, to fit model parameters to field observations. Their findings show that two factors that critically affect estimates of water in minerals were not well defined in the previous models: one is the presence of excess charges (protons) that diffuse through the network of water molecules by a process called proton hopping, and the second is the amount by which water in mantle minerals reduces its total energy or enthalpy.At present, the lack of petrological observations of the amount of water in mantle minerals at most sites in the world makes it difficult to constrain the above two factors. On the basis of their study, the authors propose more petrological studies and calibration of laboratory models with other geophysical and petrological field observations that may prove invaluable in understanding the behavior of the mantle.

Abstract | Full Text | PDF