Frost {GCDkit} | R Documentation |
Classification of granitic rocks proposed by Frost et al (2001).
Frost(plot.txt = getOption("gcd.plot.text"), clssf = FALSE, GUI = FALSE)
plot.txt |
logical, annotate fields by their names? |
clssf |
logical, should the samples be classified? |
GUI |
logical, is the function called from a GUI? |
Classification scheme proposed by Frost et al. (2001). It consists of three diagrams:
Fe number vs. SiO2 (in wt. %). NB that the Fe-number is calculated as weight proportion of FeO/(FeO+MgO) (or FeOtot/(FeOtot+MgO)). See also Warning below.
SiO2 (in wt. %) vs. Modified alkali-lime index (MALI), defined as Na2O+K2O-CaO (in wt. %).
SiO2 (in wt. %) vs. Aluminium saturation index (ASI), defined as molecular Al2O3/(Na2O+K2O+CaO-1.67 P2O5).
In fact, this is the A/CNK parameter of Shand (1943), corrected for the Ca content in apatite. See also Warning below.
The classification is designed to work both with analyses distinguishing between ferrous and ferric iron (preferred) and those with total iron only. The dialogue box lets the user decide, whether to use the ferrous iron value or the total iron.
Similarly, if some P2O5 concentrations are missing in the dataset, the user is prompted whether the missing values should be replaced with zero. If not, the problematic analyses are not plotted/classified.
The following associations are distinguished:
ferroan, magnesian
alkalic, alkali-calcic, calc-alkalic, calcic
peralkaline, metaluminous, peraluminous
The geologically reasonable combinations, together with examples, are listed in the ../Diagrams/Geotectonic/FrostTable.html, modified from the original article.
In fact, the ASI vs. A/NK diagram (based on Shand (1943)) was not plotted in the paper, but it replaces the conditions mentioned in the text and is, in our view, more instructive.
results |
The function returns table of calculated coefficients (Fe-Number, MALI, ASI). |
There are two values for the ASI: that labeled 'ASI'
is calculated from molecular
proportions of oxides, and is used for plotting and classification.
The other one is labeled 'ASI_orig'
, and is calculated exactly as stated
in the original paper (i.e. Al/(Ca-1.67P+Na+K).
This function uses the plates concept. The individual plots can be selected and their properties/appearance changed as if they were stand alone Figaro-compatible plots.
See Plate
,
Plate editing
and
figaro
for details.
Due to the specific design of this scheme (combination of multiple diagrams),
the classification option is not available via the pull-down menus.
Currently, the only way to apply Frost's classification in GCDkit on
individual samples is to call the function manually
from the Console: Frost_2008(clssf = TRUE)
.
Note that the Fe-number is calculated as weight proportion of FeO/(FeO+MgO) (or FeOtot/(FeOtot+MgO)). The approach used here should not be confused with the more common usage of the term "Fe-number" (as well as "Mg-number") for molecular proportions.
As approved by one of the authors (C. Barnes, pers. comm., 2008), the equation for ASI in the original work (Frost et al. 2001) was stated erroneously as molecular proportions of elements, instead of oxides.
This function uses the plates concept. The individual plots can be selected and their
properties/appearance changed as if they were stand alone Figaro-compatible plots.
See Plate
,
Plate editing
and
figaro
for details.
Vojtěch Erban, erban@sopky.cz
& Vojtěch Janoušek vojtech.janousek@geology.cz
Frost BR, Barnes CG, Collins WJ, Arculus RJ, Ellis DJ, Frost CD (2001) A geochemical classification for granitic rocks. J Petrol 42: 2033-2048. doi: 10.1093/petrology/42.11.2033
Shand SJ (1943) Eruptive rocks, 2nd ed. John Wiley, New York, pp 1-444
Frost_2008 Shand classify Plate Plate editing plotPlate figaro
data(sazava) accessVar("sazava") # Plot the diagrams plotPlate("Frost") # Classify the samples, suppress the graphical output Frost(clssf=TRUE, GUI=TRUE)