Wednesday, March 05, 2008


A number of geobloggers have been posting cool looking outcrop or deskcrop pictures with the aim of letting us guess what they are. As a geochemist, I find this rather passé. Firstly, the more interesting a rock looks, the less likely it is to be useful for geochemical extrapolation. Geochemists prefer featureless, homogenized rocks, and if necessary, we will homogenize the pretty rocks by grinding them into powder. One we make a rock look boring, we generally then measure the chemical composition of that rock, and use our measurements to make vast extrapolations about the history of our rock, our planet, and even the formation of the solar system. And this all comes from squiggly graphs.

So, as an inaugural geochemical puzzler, I present the following.

These squiggles show the concentration of the Rare Earth Elements (known to non-geologists as lanthanides) in rocks. The aim of this contest is to guess what rocks they are from, and why they are important. Bonus points are awarded for telling us why taking these squiggles for granted can lead to all sorts of trouble. And anyone who can rattle off the references these are from just by looking at the graphs wins the title of geochemoblogpspherohero.


Anonymous said...

Well, there appears to be an enrichment in the LREE, so we're not looking at a rock formed through volatile evaporation.

We have enrichments in even-numbered elements, which if I recall correctly, is how things go from nucleosynthesis in stars. With a preference in even-numbered elements over odd-numbered elements.

If the ppb are correct, then it's hitting really close to home for me, because it looks, apart from the Pm and Sm measurement, like the REE pattern from a bulk CI1 chondrite.

Did I win?

Anonymous said...

I should probably expand on that. CI chondrites have very similar concentrations of certain refractory and other elements (Al, Ca, Ti, V, Sr, Y, Zr, Nb, Ba, La, Ce, Pr, Nd, Sm, Eu, Gd, Dy, Er, Lu) where the Sun/CI ratio is 1.004 ±0.12, so they make good proxies for the non-volatile composition of the solar nebula. Because of this, CI chondrites are used as a normalising standard for a lot of planetary geochemistry (including mantle geochem). You need to take special care in the squiggles, because a several ppb change in one element over another involving processes working over 4.5 billion years, can cause big changes and tell you a lot about the rock's history (especially when you've got fractional crystalisation and the like happening over and over again).

I think the red line graph is probably Pulme and Beer (1993) or Lodders (2003), but I'm guessing. The higher blue line is above any values I've used, so I'm stuck there.

Ron Schott said...

Nice illustration of the Oddo-Harkins effect. Since it's standard practice to plot REE in chondrite-normalized abundances, I'd have to guess you're beginning this geochemical lesson with a basic chondrite, as Chris suggests. I'll check my references when I get in to work later...

Ron Schott said...

It looks like your red/purple data may come from Anders and Grevesse (1988) which is what I used for chondrite normalization last time I worked with REE data. Not sure where the blue data points come from.

C W Magee said...

Both of y'all are mostly right.

The blue line is an ordinary chondrite. The other line is actually 4 lines- All different C1 chondrite reported measurements. One of which is Anders & Grevesse '89. Details tonight.