Dug,
You tell 'em! I do agree with both Dave and you. We need better
thermodynamic data and mixing models for metamorphic cordierite, biotite,
hornblende and chlorite solid solutions. I would also argue for more
careful determinations of Fe3+/Fe2+ and O/OH in hydrous ferromagnesian
silicates. The Fe3+/Fe2+ may also be reset along with O/OH during late
loss/gain of H. Cordierite may not preserve its peak H2O content (once
determined). Of course, one has to address disequilibria, esp. in rocks
that are rapidly heated or slowly cooled.
Eric
Dave wrote:
>>... Currently
>>existing thermodynamic data sets do a moderate to poor job of handling
>>natural low pressure mineral assemblage sequences. Rectification of this
>>deficiency is a worthwhile goal, because variations in mineral assemblages
>>are far more useful in providing information on subtle P-T variations in
>>low-P (And-Sil type) settings than geothermobarometry - see my chapter on
>>metapelites in the Am Min 'Contact metamorphism' volume (v. 26, 1991) to
>>see why.
>
Dugald's comment
>I have reread this chapter without seeing why. The accuracy of both the
>grid and the thermobarometers depends on the same thing - accurate
>determination of the thermodynamic properties of minerals and their
>inferred-to-have-been-coexisting fluid and/or liquid. The so-called
>"constraints of Schreinemakers' analysis" (op cit p.109) are nothing more
>than the constraints of stable equilibrium. It is not possible to "rectify"
>the grid without simultaneously making it possible to rectify the
>thermobarometers. But even if/when you youngsters have perfectly accurate
>thermobarometers and perfectly accurate grids for the system TiKNFMASH at
>specified mol fractions of Mn in Grt, Ca in Plg, Zn in St, Fe3+ in Ilm
>(etc.), and even if/when you can make perfectly accurate mineral analyses,
>there will still be no certain way to prove that "geothermobarometric
>results... which are inconsistent with the observed phase relations" (op cit
>p.109) are due to metastable mineral compositions rather than unstable phase
>relations.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|