of seamounts, islands and plateaus on the Pacific plate
2.1 - 15 May 2004
Clouard1 & Alain Bonneville2
de Geofisica, U. de Chile, Blanco Encalada 2002,Casilla
2777 ,Santiago, Chile, firstname.lastname@example.org
Géosciences Marines, Institut de Physique du
Globe, CNRS, 4, Place Jussieu,75252 Paris Cedex 05,
|Click here to go to Discussion
of this page
The report made available here presents a compilation of reliable dating
of seamounts and islands on the Pacific plate from 1,300 samples. Paleomagnetic
ages from seamount magnetism are not included. Furthermore, we do not
consider foraminifer ages which only give the minimum ages of seamounts.
Only radiometric ages are included, and Ar-Ar ages are preferred over
K-Ar ages when both are available from the same location.
The table gives our compilation along with references to the selected
data. An average age [1, 2] is given when several ages have been determined
for the same volcanic stage. In this case the geographical coordinates
are given only once for the line. A similar procedure is followed when
several volcanic stages exist for the same seamount or island. NB: an
ASCII file (ages_pacific_v2.1.dat) with only average ages and geographical
coordinates is also available for plotting purposes.
Click here to download the average
ages and geographical coordinates (Excel spreadsheet).
Click here to download the article “Ages
of seamounts, islands and plateaus on the Pacific plate”
- Clouard, V., Etude géodynamique et structurale du volcanisme
de la Polynésie française de 84 Ma à l'actuel,
Ph.D., Université de la Polynésie française, 2000.
- Clouard, V. and A. Bonneville, How many Pacific hotspots are fed
by deep-mantle plumes?, Geology, 21, 695-698,
Wed July 21st, 2004: Ajoy Baksi
A look at this effort shows that data from over 50 different papers were
examined. It would appear that the original authors assertions as to the
reliability of ages, have been accepted without question. Looking at some
of the references, I know that some of the ages therein are totally untrustworthy,
(almost) as bad as the ones I took up in my page (see ArAr.html).
What is the criterion for a reliable age? Should the
radiometrically determined “number” gauge the crystallization
time within (say) ± 10% or should it be to within ± 2% (1-sigma
estimate)? The latter should be the preferred way for purposes of “tracking
hotspots” properly; except for the youngest ages, where high atmospheric
contamination levels would produce large analytical errors. However, I
note that a 40Ar/39Ar determination of the age of
the most recent geomagnetic reversal (Matuyama to Brunhes), gave an age
of 783 ka with an error estimate of less than 2% (Baksi et al.,
Science, 1992). It is clear that in most cases, K-Ar dates cannot be taken
to be accurate estimates of crystallization age for whole-rock basalts.
Thus the careful K-Ar work (a large part on whole-rock basalts), setting
up the geomagnetic polarity time scale for 5-1 Ma, was shown to have systematic
errors averaging 5-7% (Baksi, GRL, 1995), resulting from alteration
and partial loss of radiogenic argon (40Ar*). This was for
samples critically screened for freshness by thin section examination!
40Ar/39Ar total fusion “ages” should
not normally be taken as accurate estimates of crystallization age (see
Baksi, Jour. Geol., 1999). We are then left only with 40Ar/39Ar
stepheating results; and all authors have not been equally scrupulous
in interpreting plateau and isochron data (see Baksi, 1999, and
my webpage on this site: ArAr.html).
Many authors have been very careful in interpreting 40Ar/39Ar
data, e.g. Brent Dalrymple and Marvin Lanphere. I wouldn’t be surprised
to see that in cases, what they may have tentatively estimated to be a
crystallization age, has now become entrenched in the literature as “reliable”
or “very reliable”. Argon dating of altered whole-rock mafic
material – as with almost all of the samples from “Pacific
Seamounts” - is always a tricky business. On what basis is one to
assess their freshness? Is it possible to gauge this from the 40Ar/39Ar
raw data? I believe this may be possible, and I am working on some manuscripts.
My suspicion is that on critical examination, a number of “accepted”
ages for Pacific Seamounts may bite the dust; my fear is that it may hit
epidemic proportions, with > 50% of the “ages” having to