No Plumes Along Mid-Ocean Ridges


Presnall (2008)
No Plumes Along Mid-Ocean Ridges
Dean C. Presnall
Department of Geosciences, University of Texas at Dallas, Texas, USA
Bayerisches Geoinstitut, Universität Bayreuth, Germany
Geophysical Laboratory, Washington, D. C., USA
Hot plumes were considered by Morgan (1971, 1972) to be basic components of a whole-mantle
thermal convection regime that guides and helps maintain volcanism along spreading ridges. These
plumes have been considered to be the main or even total (Yamamoto et al., 2007) source of thermal
energy that drives plates. They would also provide the energy for enhanced lava production and
locally higher temperatures along spreading ridge. A recent paper (Presnall & Gudfinnsson, 2007)
presents a new model for the generation of mid-ocean ridge basalts (MORBs) and oceanic lithosphere
formation. This model addresses the temperatures of melt extraction along mid-ocean ridges, an
approach toward mapping of mantle heterogeneity beneath ridges, and the existence of hot plumes at
volcanic centers (e.g., Jan Mayen, Iceland, Azores, St. Helena, Tristan, Bouvet, Afar, Easter) on or
close to ridges.
An earlier model (Klein & Langmuir, 1987; 1989; Langmuir et al., 1992; hereafter LKP), involves a
range of short (cool) to long (hot) melting columns, with aggregate melts produced at the surface that
combine melt increments from the entire depth range of a column. They introduced the parameters
Na8 and Fe8 (Na
O and FeO normalized to MgO = 8%), in order to remove the variations of Na
O and
FeO caused by low-pressure fractional crystallization and to reveal the chemical systematics produced
by melting. To explain variations of Na8 and Fe8 vs. axial ridge depth, LKP claimed that long melting
columns that start at high T and P produce melts that are high in Fe8, low in Na8, and characteristic of
shallow ridges with a thick crust. In contrast, they proposed that short melting columns that start at low
T and P produce melts that are low in Fe8, high in Na8, and characteristic of deep ridges with a thin
crust. These were called the global trends and were considered to be characteristic of ridge lengths of
hundreds to thousands of km.
In a global reevaluation of the LKP modeling, Presnall & Gudfinnsson (2007) confirmed that when Na8
and Fe8 are plotted against each other, rough positive and inverse trends occur, as reported by LKP.
However, no ridge segments with “global” lengths were found that support an increase of Na8 and
decrease of Fe8 with increasing axial ridge depth. This absence of this key correlation requires a
different kind of modeling. In the system CaO-MgO-Al
O-FeO (CMASNF), solidus melt
compositions in the plagioclase/spinel lherzolite transition (~0.9-1.5 GPa) show an inverse correlation
of Na
O vs. FeO at constant MgO (Presnall et al., 2002; Presnall & Gudfinnsson, 2007). This matches
the global Na8-Fe8 correlation of LKP but occurs for solidus melts over a very narrow P-T range of
~1.2-1.5 GPa and ~1250-1280°C.
Also, if MgO is relatively uniform but allowed to vary slightly, Na
O varies positively with FeO during
progressive fractional melting of ascending mantle source material (Presnall & Gudfinnsson, 2007). In
this modeling, the inverse variation of Na8 vs. Fe8 is caused by mantle heterogeneity rather than
temperature variations. The positive correlation is produced by fractional melting over a narrow
pressure range. For example, Figure 1a shows, for the Carlsberg Ridge, a ridge of “global” length


00 votes

Related Articles

Related forms

Related Categories

Parent category: Miscellaneous
Page of 4