Lateral melt corner flow in the lower ocean crust
Abstract
The gabbroic layer of the oceanic crust is formed by the crystallization of basaltic melt, but the mineralogy of the gabbroic layer as sampled at the ocean floor doesn't add up: It is far too evolved. For example: A primitive MORB liquid is predicted (using MELTS) to crystallize twice as much dunite (10%) and troctolite (12%) together as oxide bearing gabbro (12%), yet cumulate dunites are essentially not found and cumulate troctolites are rare, while oxide bearing gabbros are very common, forming about a majority of the total section near the transform at Atlantis Bank and Atlantis Massif, to name two prominent crustal sections.
We normally explain this away by saying the troctolites are present in the section but not yet drilled. Yet, in areas of lower crustal outcrop we can readily sample mantle rocks at the sea floor, suggesting that these deeper cumulates are simply not present in the section we sample. This has been called the "missing troctolite problem". Part of the problem is simple sampling bias: the huge majority of sampling sites for lower crustal rocks are at segment ends, which may not represent the entire crustal section. If this is so, then lateral melt transport in the lower crust is required to produce the observed outcrop patterns. In a lateral flow model, the lower crustal accretion is dunitic/troctolitic at its center, undergoing lateral flow to the segment ends. The melt lens and basaltic section sit on top of this forming an essentially separate surficial magmatic plumbing system in the brittle regime. This model would have several interesting implications. First, at any given point, there is no vertical mass balance. That is, a vertical hole drilled at the segment center would retrieve relatively too few evolved gabbros beneath the volcanic complex, instead becoming troctolitic fairly rapidly. At the one place where the lower crust could be drilled away from a major transform fault (Hess Deep), there were indeed abundant troctolites, though the section was not sufficiently complete to judge the degree of mass balance. At segment ends, the requirement for deep primitive cumulates is relaxed in this model, and nearly the whole section would be evolved oxide-bearing gabbro. In ophiolites, lateral variation from troctolites to evolved gabbros may be a simple lateral migration feature.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.T23E0424S
- Keywords:
-
- 7218 Lithosphere;
- SEISMOLOGYDE: 7220 Oceanic crust;
- SEISMOLOGYDE: 8120 Dynamics of lithosphere and mantle: general;
- TECTONOPHYSICSDE: 8416 Mid-oceanic ridge processes;
- VOLCANOLOGY