Ontong Java Plateau – the Largest
and the Most Puzzling
of Geology and Geophysics, Yale University, P.O.
Box 208109, New Haven, CT 06520-8109
For discussion of this
page, click here.
This short essay is based
on my recently published paper:
J., Why did not the Ontong Java Plateau form subaerially?,
Earth Planet. Sci. Lett., 234,
Please read this paper
for more details. Here, I emphasize two issues in the
broader context of currently ongoing debates on the
existence of mantle plumes.
Figure 1. Location
of the Ontong Java Plateau with previous drilling sites.
4000-m bathymetric contour is shown as a line to indicate
the spatial extent of this plateau.
First, it is essential
to consider ALL available data. Of course, not all data
are of the same quality and reliability, so one always
has to be careful to not weight equally strong and weak
constraints. Nevertheless, it is usually beneficial
to look at something from different perspectives (e.g.,
geochemical vs. geophysical). The generation of hotspots
and large igneous provinces involves chemical differentiation
due to mantle melting, which has both geochemical and
geophysical aspects. If one's working hypothesis is
entirely geochemical or entirely geophysical, one may
miss something critical.
Second, it is important
to realize that we are all limited by our imaginations
when searching for the truth with incomplete data. We
may not be able to reach conclusions through a process
of elimination. In the case of the Ontong Java Plateau,
for example, at the time I wrote my recent paper (Korenaga,
2005), there were two major hypotheses: the plume
hypothesis and the bolide impact hypothesis (Ed: See
vs. Alternative Hypotheses for the Origin of the Ontong
Java Plateau and Impact
Origin for the Greater Ontong Java Plateau?). The
plume hypothesis, at least in its classical form, cannot
explain the submarine eruption of the Ontong Java Plateau
nor its anomalous subsidence history (Ed: See also Sedimentary
Evidence for Moderate Mantle Temperature Anomalies Associated
with Hotspot Volcanism). But this does not mean
that the bolide impact hypothesis is more likely. As
it turns out, the bolide impact hypothesis is even less
likely than the plume hypothesis, given the currently
available geophysical and geochemical observations.
Plume vs. non-plume debates
tend to be formulated as competition between two contrasting
hypotheses, but there may be many other possibilities.
Consider linear differential equations vs. nonlinear
differential equations. The latter encompass everything
that is not linear, so in a sense, they are not well
defined. This may frustrate discussion because non-plume
explanations such as chemically heterogeneous mantle
may appear to be ad hoc; one can choose source
composition to be just right to explain many observations,
if not everything. However, chemical heterogeneity is
a plausible idea with growing observational support.
It is thus worth putting effort in to understand better
the melting of heterogeneous media and resulting mantle
dynamics. The interpretation of observations depends
heavily on our theoretical understanding.
With these two points
in mind, I reiterate the main conclusions of my recent
paper on the Ontong Java Plateau (Korenaga,
2005). Key observations to be considered are the
- The bulk of the plateau formed at ~120
Ma (less than a few million years).
- The plateau was formed on relatively
young seafloor (15-30 Ma).
- Average crustal thickness is greater
than 30 km (the total crustal volume for OJP is estimated
to be greater than 40-50 x 106 km3).
- The bulk of the plateau was formed
in a submarine environment, probably more than 1 km
below sea level.
- The plateau has subsided by only ~1-2
km since its formation.
- There was minor magmatism at ~90 Ma.
- The plateau is underlain by unusually
low seismic velocities.
- Most of the erupted lava is isotopically
very uniform over a vast area (~1000 km).
The first three observations
are fairly robust, and also not very surprising just
by themselves. Other oceanic plateaus in the Pacific
are similar in these respects. However, the rest of
the observations listed are extremely difficult to explain
if the first three are correct. The (thermal) plume
hypothesis cannot explain observations 4-7, and probably
not observation 8, if one considers that a convecting
plume head is expected to have entrained ambient heterogeneous
mantle. The bolide impact hypothesis also cannot explain
observations 4-7, and probably not observation 8 either.
More seriously, the impact hypothesis does not explain
observation 3, which is fatal (Ed: See also
Impact volcanism as a possible origin for the Ontong
Java Plateau (OJP)). The alternative hypothesis
I proposed in my paper, which calls for the entrainment
of dense recycled oceanic crust by rapid seafloor spreading,
can potentially explain observations 1-6. It is still
not clear to me how to explain observations 7 and 8.
As far as I know, none of the (now) three existing models
can explain observation 7 in a convincing manner. The
first thing to do may be to study in more detail the
mantle velocity structure and its uncertainty with new
The origin of the Ontong
Java Plateau is thus still not completely resolved.
Even though this plateau is, to date, the best sampled
oceanic plateau, we can still learn much more by collecting
new data, both geochemical and geophysical. Equally
importantly, we need to improve our understanding of
the mantle dynamics associated with the melting of heterogeneous
sources (Ed: See also The
layered mantle revisited and The
eclogite engine: Chemical geodynamics as a Galileo thermometer).
The Ontong Java Plateau may have been formed by some
kind of plume, but if so, it must have been very different
from what we envisage a mantle plume to be. New data
acquisition, better theoretical understanding, and feedback
between observation and theory are the best ways of
making further progress.
25th April, 2006, Stephanie Ingle
Several of the conclusions listed in this webpage have
been known for quite a while. It might be helpful to
bring to the readers' attention some relevant references:
- ..bulk of the plateau formed at ~120
References: Mahoney et al., Geophys. Monogr.
77, 1993; Tejada et al., J. Petrol.,
- ..the plateau formed on ... young
References: Andrews et al., DSDP Init. Reports
30, 1975; Sliter & Leckie, ODP
Init. Reports Leg 130, 1993; Mahoney et al.,
Proc. ODP Init. Reports 192, 2001; Ingle
& Coffin, EPSL, 2004.
- Average crustal thickness...
References: Gladczenko et al., JGR, 1997;
Richardson et al., Phys. Earth Planet. Interior,
- .. plateau formed in a submarine environment...
References: ODP Leg 192 Shipboard Scientific Party,
2001; Tejada et al., J. Petrol. 2002.
- Plateau has subsided only 1-2 km since
References: Andrews et al., DSDP 30,
1975; Sliter & Leckie, ODP Init. Reports Leg
130, 1993; Shipboard Scientific Party, ODP
Leg 192, 2001; Ingle & Coffin, 2004; Roberge
et al., Geology, 2004.
- Minor phase of magmatism at ~90 Ma.
References: Tejada et al., J. Petrol. 1998;
- ... unusually low seismic velocities.
Reference: Richardson et al., Phys. Earth Planet.
- .. isotopically very uniform...
References: Mahoney, Geophys. Monogr. 43,
1987; Castillo et al., EPSL 1991; Castillo
et al., Proc. ODP, Sci. Res. 129,
1992; Castillo et al., EPSL, 1994; Tejada
et al., J. Petrol., 1998, 2002. [and others]
25th April, 2006, Jun Korenaga
Thanks much, Stephanie, for listing up relevant references,
most of which are, of course, cited in my EPSL paper.
Importantly, however, the list of key observations is
not my own "conclusion" list – it's
just a list of observations. What follows after
the list is the conclusion of my paper.
last updated 2nd August, 2020