Initial Publication Date: October 19, 2006
Part 4 - Prepare the Data for Graphing
Check for rows of anomalous data
At this point, it is a good idea to check the data for rows that you do not want to include. Remember that GEOROC datasets include all data published, including partial analyses, sedimentary and metamorphic rocks, xenoliths, and altered samples. For altered rocks, it is probably a good idea to delete 'extensively-altered' (E) and 'moderately-altered' (M) samples. Altered samples may or may not be present in your downloaded dataset.Why don't we want to use altered samples?
If we were interested in the geochemical effects that hydrothermal fluids have on rocks, we would definitely want to include altered samples in our dataset. But for this exercise, we are more interested in how rock samples can be used to determine something about the rock when it was initially forming at high-temperature (magmatic) conditions. This can only be done if the rock sample has been more or less a closed-system since its formation with respect to the various geochemical components for which we have compositional data. Alteration usually involves leaching of components out of the rock, and/or the precipiation of new minerals within the rock.
If we were interested in the geochemical effects that hydrothermal fluids have on rocks, we would definitely want to include altered samples in our dataset. But for this exercise, we are more interested in how rock samples can be used to determine something about the rock when it was initially forming at high-temperature (magmatic) conditions. This can only be done if the rock sample has been more or less a closed-system since its formation with respect to the various geochemical components for which we have compositional data. Alteration usually involves leaching of components out of the rock, and/or the precipiation of new minerals within the rock.
Recast Fe oxide data as FeOT
GEOROC does not have a consistent way of reporting whole-rock iron oxide contents, so you have to go through each spreadsheet and make this change manually. 'FeOT' is probably the best way to go. For samples containing Fe-oxide data but lacking a value for FeOT, insert this formula to calculate it: = Fe2O3T*0.8998 + Fe2O3*0.8998 + FeO. Do not copy this formula to the rows lacking any Fe-oxide data.Actually, its not that simple
When we convert Fe2O3 into FeO, there is a net change (decrease) in the number of oxygen atoms on a per-Fe basis. You knew this because Fe2+ (in FeO) is the reduced cation, and Fe3+ (in Fe2O3) is the oxidized cation. This decrease in oxygen atoms is equivalent to decreasing the analysis total by a small amount. If we were rigorous about this, we would want to re-normalize all the element oxides to the smaller total, but since this renormalization results in a only small changes for the various element oxides, we'll just ignore it for this exericse.
When we convert Fe2O3 into FeO, there is a net change (decrease) in the number of oxygen atoms on a per-Fe basis. You knew this because Fe2+ (in FeO) is the reduced cation, and Fe3+ (in Fe2O3) is the oxidized cation. This decrease in oxygen atoms is equivalent to decreasing the analysis total by a small amount. If we were rigorous about this, we would want to re-normalize all the element oxides to the smaller total, but since this renormalization results in a only small changes for the various element oxides, we'll just ignore it for this exericse.