Is the D'' recording present and past lower mantle melting?
Abstract
Seismic measurements show that generally the D'' is a layer about 200 km thick, characterized by anomalous, mostly high, seismic velocities. In contrast, in some regions, a thin layer of ultra low seismic velocities (ULV) is observed. The ULV zones agree well with a scenario of up to 30 percent partial melt underneath plumes (Williams and Garnero 96). I investigate the possibility that the ULV zone is a region of lower mantle partial melting, where the melt produced is denser than the residual matrix. In this scenario heavy melts drain downwards, while the depleted, lighter matrix upwells in mantle plumes. Such melting, requiring heat supplied by the core, will occur where the lower mantle solidus cuts the local lower mantle temperature profile. Recent experiments on melting of lower mantle material (Boehler 2000) suggest that indeed the lower mantle solidus may cut the lower mantle temperature profile close to the CMB. However, since the earth is cooling over time, if melting is a possibility now, it was certainly even more probable when the Earth was younger and hotter. Thus it is expected that the point where the lower mantle adiabat cuts the solidus is becoming progressively deeper with time, and the thickness of the partially molten zone is progressively shrinking. I investigate the possibility that the D" as a whole is recording an ancient, now mostly crystallized, sea-level of lower mantle partial melt, where the currently observed ULV zones are merely residuals from much more glorious days of lower mantle partial melting.
- Publication:
-
EGS - AGU - EUG Joint Assembly
- Pub Date:
- April 2003
- Bibcode:
- 2003EAEJA.....8233A