Geology 423 – Problem 5:
Trace Element Modeling of Igneous Processes
The attached data set includes trace element analyses of some Crator-Lake, Oregon, calc-alkaline rocks. Use these analyses plus the attached list of trace element Distribution Coefficients to model the trace element differentiation at Crator Lake.
1. Use the attached phase diagram to estimate the relative plagioclase, augite, and olivine abundances in your model (assume that all three are in equilibrium with the magma). Use the included elements: Sc, Cr, Co, & Ni, and the excluded elements Rb, Y, Zr, & Cs, to calculate the expected trace element contents of daughter magmas for 5, 10, 20, and 40% crystallization assuming both Nernst fractionation and Rayleigh fractionation (i.e.- you will have two models) and using the Crator Lake Basalt as your parent magma. If you set this up on an excel spreadsheet, once you have calculated the results for one element, you can copy the equations, change the Kds, and it will instantly calculate the other elements.
2. How do the Nernst and Rayleigh models compare? How do your calculated results compare with what is observed in the data set? What element(s) can you use to determine the crystallization %s in the data set? (Hint – You should use the most excluded element to calculate the % crystallization) Does it look like any of your three minerals are over-represented in your model? Does it look like any of the KDs are distinctly off?
3. Calculate the expected REE patterns for your Nernst and Rayleigh models after 5, 10, 20, and 40% crystallization. Prepare plots of the normalized REEs (normalize by dividing the sample concentration by the concentration in chondrites) as a function of ionic radius. To get a good model you should extrapolate the Kd values for the REEs assuming that they will change in a regular fashion as a function of ionic radius with the exception of Eu. Remember there is no Pm in your samples as it is not a naturally occurring…