Dear Doug
Your point re albite and the garnet equilibrium domain takes us to the heart
of one of the things that I would like to examine in detail at discussion
sessions that I would love to have us run at the conference.
By the way, we got no more abstract offers from my appeal! Not one!
Perhaps some of you will just come and take part in structured and led
discussion sessions that we could run when economic or structural stuff is
on that no-one is interested in. The list of all conference talks will be
available soon.
To set the scene we have been determining the FIAs to get the succession of
garnet porphs resolved and then probing the cores of (these days) mainly
single FIA samples.
1. We commonly, for non-calc pelites, get very tight Mn Ca and Fe isopleth
intersections for the core on a PT pseudosection that lie on or very close
to the garnet in line. Indeed, the results are TOO GOOD!!! I suspect that in
many cases the PT of nucleation was below the garnet in line, but as this is
impossible in THERMOCALC, it was ³snapped² up to the garnet in line by the
program. This requires either a bulk compositional shift (e.g., dissolution
of plag if some was or is present), or that THERMOCALC is not as accurate as
one might wish.
2. For calc pelites we can get quite tight Mn, Ca, Fe intersections that lie
well above, sometimes large amounts above, the garnet in line. This suggests
a bulk composition shift due to cleavage development post porph growth.
However, another possibility is that the rocks had reached higher PT than
the garnet in line but did not grow garnet until deformation partitioned
through the outcrop at the scale of a porphyroblast. HERESY!!! NO! We have
proven that this possibility occurs structurally and we are working at
getting metamorphic data that will convince you via 1 above and 3 below.
3. This leads us to the main problem. What would happen to isopleth
intersections if garnet did nucleate in a pelite a significant distance in
PT above the garnet in line because no deformation partitioned through that
out crop at that scale until the PT were well and truly overstepped for that
bulk composition. Would isopleths still intersect in a point (we have
examples of this or is this simply a problem with THERMOCALC accuracy?).
Alternatively, if they could not intersect at a point, just what would
happen. One of us set out to look at field gradients using FIAs and
isopleths. He selected 12 samples containing the 2nd to youngest FIA in
garnet cores (from a succession of 6) and another 12 samples containing the
3rd to youngest FIA in garnet cores, in a region where we know the T&P went
up throughout the FIA history. This means that some of these samples
compositionally might have been able to grow garnet at an earlier stage in
the deformation and metamorphic history, but did not because no deformation
partitioned through that outcrop at the scale of a porphyroblast until late
in the regional deformation and metamorphic history. After a large amount
of effort, decent isopleth intersections still have yet to be obtained for
these samples. Why? For several, Ca levels are low in the bulk composition.
Some plagioclase may have been present that was dissolved post garnet growth
as the matrix foliation intensified. Alternatively, or as well, could the Ca
core isopleth be significantly shifted by the effects of dramatic
overstepping, such that an intersection that even remotely approaches the
Fe, Mn core isopleth intersection is simply not possible. Some here have
run into this Ca isopleth problem previously and simply given up on the
sample and moved on to others. If we have this problem then met pets
elsewhere have run into it as well. I am convinced that discarding such
samples for whatever reason, or putting them in the too hard basket and
moving on, and then not discussing them in the literature is dramatically
hindering advancement in understanding. Instead of neglecting to talk about
the problem samples, we need to see if we can test whether the structural
timing control on porphyroblast growth that FIAs provide, allows us to
1. negate at least some of the effect of compositional shifts,
2. determine the level of accuracy of THERMOCALC
3. test whether the structurally indicated solution of overstepping
provides the answer
4. test any role for the scale of the equilibrium domain being say only
twice the size of a porphyroblast (there are microstructural reasons for
this possibility).
5. and develop the concepts needed to allow significant advancement in
understanding the processes that are taking place!
We hope to have made enough progress on this one by the conference to be
able to let you know about number 3 above. Any suggestions for what you
think conceptually ought to take place with isopleths would be very welcome.
Any suggestions for what we can do to test this are welcome.
If enough samples give tight core isopleth intersections to pin down the TP
range for each of the 2 FIAs mentioned above, that researcher may be able to
work backwards to resolving the ones where the isopleth intersections are
problematical. I wonder what others have found with such samples.
Cheers
Tim
> At 12:45 PM 07/06/2005 +1000, you wrote:
>> Eric
>> We deliberately do not work with calc pelites, for this type of work, to
>> avoid exactly the problems that you mention. Every time a differentiated
>> crenulated cleavage forms (this occurs in most deformations and, indeed,
>> shuts down porphyroblast growth -if you have any queries about that read
>> Bell & Hayward, 1991 (JMG) and Bell et al 2004 (JSG) first and then fire
>> away), there is significant volume loss due to the dissolution of non platy
>> or linear mineral phases, which shifts the bulk composition from the time
>> when the porphyroblasts grew.
>
> Hi Tim,
>
> What I write below is meant to be restricted to using phase diagram
> sections for determining garnet-growth conditions.
>
> These are very important points you make; it has caused some headaches for
> me. However, dissolution-mobilization-regrowth of plagioclase can also be a
> blessing in disguise in some cases, can't it. If the rate of plagioclase
> growth = rate of plagioclase dissolution, and
> Grt_porph_diameter/compositional_layering_width is high, then plagioclase
> dissolution-regrowth may actually be required in order to make approaching
> equilibrium a possibility (in uppermost greenschist-amphibolite facies at
> least). Of course you are right about significant volume loss, if the
> volume loss is measured with respect to a hypothetical chemical equilibrium
> domain. The problem is knowing what that garnet equilibrium domain is.
>
> Regarding steering clear of calc-pelites: do you define any pelite with
> plagioclase a calc-pelite? If you steer clear of plagioclase-bearing
> pelites, how is one to be certain that the reason the pelite does not
> currently have plagioclase is because of the reasons you describe
> (plagioclase component volume loss)? It seems that in rocks that many would
> call a "normal pelite", plagioclase/albite actually is a phase present at
> garnet nucleation T's at moderate-low crustal P's. I sometimes worry about
> working with pelites that don't have plagioclase, for fear of the volume
> loss you describe.
>
> Cheers,
> Doug
>
>
>
>> We certainly take into account Mn. Our pseudosections are most commonly
>> calculated in the NCMnKFMASH system.
>>
>> Cheers
>> Tim
>>
>> Prof. T.H. Bell
>> School Earth Sciences
>> James Cook University
>> Townsville
>> Qld 4811
>> Australia
>> Work Phone +61-7-47814766
>> Work Fax +61-7-47251501
>> Home Phone+61-7-47732534
>> Email [log in to unmask]
>> http://www.es.jcu.edu.au/dept/Earth/research/samri/index.html
>
> Doug Tinkham
> Dept. of Geology and Geophysics
> University of Calgary
> 2500 University Drive NW
> Calgary, Alberta T2N-1N4 Canada
> Ph:403-220-8209
> e-mail: [log in to unmask]
Prof. T.H. Bell
School Earth Sciences
James Cook University
Townsville
Qld 4811
Australia
Work Phone +61-7-47814766
Work Fax +61-7-47251501
Home Phone+61-7-47732534
Email [log in to unmask]
http://www.es.jcu.edu.au/dept/Earth/research/samri/index.html
|
