Abstract

Various methods for inferring the water distribution in Earth's mantle are reviewed including geochemical and geophysical methods. The geochemical approach using the water contents of basalts shows that the water content in the source regions of ocean island basalt is generally larger than that of the source region of mid-ocean ridge basalt, but the location of the source regions of ocean island basalts is poorly constrained. Geophysical approaches have potential of providing constraints on the spatial distribution of water but their usefulness depends critically on the sensitivity of geophysical observations to water content relative to other factors, in addition to the resolution of geophysical observations. Existing experimental data on the influence of water on seismologically observable properties and on electrical conductivity are reviewed. Frequently used seismological observations such as the anomalies in seismic wave velocities and of the topography on the mantle discontinuities are only weakly sensitive to water content but more sensitive to other factors such as the major element chemistry and temperature for a typical range of water contents. In contrast, electrical conductivity is highly sensitive to water content and only modestly sensitive to other factors such as temperature, oxygen fugacity and major element chemistry. Models of electrical conductivity–depth profiles are constructed where the influence of hydrogen and iron partitioning among coexisting minerals and of the depth variation in oxygen fugacity are incorporated. It is shown (i) that the electrical conductivity varies more than two orders of magnitude for a plausible range of water content in the mantle (~ 10 ppm wt to ~ 1 wt.%) and (ii) that if water content is constant with depth, there will be a drop in electrical conductivity at ~ 410-km. Although the resolution is not as high as seismological observations, geophysically inferred electrical conductivity distributions generally show higher conductivity in the mantle transition zone than the upper mantle, suggesting that the water content in the transition zone is higher than that in the upper mantle with some lateral variations. Implications of inferred water distribution are discussed including the possible partial melting near 410-km and its role in global water circulation.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call