Part 7 - Analyze and Interpret Your Results
Questions to answerYou should now have a number of variation diagrams plotted. Its time to try to make some sense out of all of this! The following are just a few question to get you thinking about what these diagrams may be telling you about the origins of the Yellowstone and Crater Lake volcanic rocks.
- Which of the two suites shows the most continuous range in composition?
- Crater Lake.
- In which location would you expect to find andesites?
- Andesites have intermediate silica contents, so the answer is Crater Lake.
- In their study of the material from the 6845 BP climactic eruption of Mount Mazama, Bacon and Druitt (1988) noted that there is a compositional gap between 61-70 wt.% SiO2 in the whole rock data. This gap is not apparent from your larger dataset which includes rocks of other ages (particularly for the SiO2 vs. MgO plot which has more datapoints), but it does seem like there are noticeably fewer samples in this compositional region on your plots. However, Bacon and Druitt note that there is no gap in the 6845 BP dataset if volcanic glasses are plotted instead of the whole-rock data. How does this fact serve to explain the origin of the compositional gap observed by Bacon and Druitt? Hint: the rocks across this compositional range are porphyritic.
- This indicates that the gap is an artifact produced by a step in the crystal content of the erupted magmas. It is not an actual gap in the liquid compositions produced under the volcano, just a gap in magma (and therefore whole-rock) composition.
- What is the term that describes the range of compositons displayed by the Yellowstone Plateau Volcanic Field?
- Based on the spread of the Yellowstone data across SiO2 (or MgO), which is the better petrogenetic hypothesis for the origin of the Yellowstone rhyolites (note: neither one may actually be true): (1) fractional crystallization of basaltic liquids compositionally similar to the basalts in the Yellowstone dataset, or (2) partial melting of mafic rocks (gabbros) similar to the basalts in the Yellowstone data?
Petrogenesis of the Yellowstone rhyolites by fractional crystallization requires a continuous range of intermediate liquids from basalt to rhyolite. For the rhyolites to have scavenged their incompatible elements (such as K and Rb) by fractional crystallization of basaltic liquids, would have required many times more volumes of basaltic liquid. The near-complete absence of intermediate rocks in the Yellowstone Plateau Volcanic Field (YPVF) and the lack of evidence of large quantitites of basalt (at least at the surface as erupted magmas) do not support this model.
A better model (but by no means proven by the major- and trace-element data) is that the rhyolites represent partial melts of mafic rocks, perhaps compositionally similar to the basalts in the YPVF. This mechanism would be consistent with the lack of intermediate magmas.
- What happens to the scatter of the data as a function of SiO2 for both the Yellowstone and Crater Lake datasets? Do you think this has petrologic significance? If so, what?
- As pointed out in Rollinson (1993) , using a dominant element-oxide such as SiO2 presents some graphical problems. One of these is a reduction in scatter with increased SiO2. This is no doubt affecting the data on your graphs. However, lower silica magmas are also lower in viscosity, so they may have experienced more crystal-liquid fractionation than higher silica magmas. This could very well be contributing to some of the scatter at low SiO2.
- What's going on with the wacko high-MgO samples from Crater Lake?
- Look for clues in the 'Sample Comment' colum in the downloaded spreadsheet.
- Several of these high-MgO samples are described as olivine scorias. Olivine is an extremely high-MgO mineral, so its a good bet that these samples contain more than their 'fair share' of olivine, and that at least some of this olivine is cumulate in origin, perhaps scavenged and concentrated by crystal-liquid fractionation (maybe settling?).
- Let's focus on the incompatible elements K, Rb, and Ba right now. What minerals commonly found in igneous rocks usually host these elements?
- Alkali feldspars, micas, and amphiboles (in order of decreasing concentration of incompatible elements).
- Describe the incompatible element (K, Rb, and Ba) trends for the Yellowstone rhyolites when plotted against SiO2 or MgO? What process(es) could be responsible for this interesting trend?
- Think about what would happen if the concentrations of these elements were solely controlled by magmatic processes involving minerals containing SiO2 and/or MgO.
- What about non-magmatic processes? Could these have affected the incompatible element concentrations?
- The trends for these three elements are almost vertical (wide variation in incompatibles vs. SiO2. Incompatible elements, especially large ion lithophiles like K, Rb, and Ba are easily partitioned into hydrothermal fluids. High-silica samples anomalously depleted in these elements may have been altered (and leached of their incompatible elements) by this mechanism. If this is true, perhaps the GEOROC dataset contains hydrothermally-altered samples which are not identified as such.