FOR SCIENCE!

*pops in to make Loud Noises about being back in a lab and being very excited*

I'm hesitent to post specifics, because I don't yet have a firm grasp of what information I'm allowed to release when. It's neat stuff, though. :>

 

Even worse, apparently it was an expensive one, so imagine them 'Borrowing' it from a parental's stache, and then whoops! *cackles*

 

What I'm doing for the next month:

 

WithAnH at the Mountains of Madness: unscientific phone pictures. I'll make a rambling post later talking in some more depth about the instruments and the actual data we're collecting, but here are a few pretty pictures of Antarctica while I'm still trying to catch up on sleep.

http://imgur.com/a/m8pmn

 

I am more amused by these maps than is probably strictly healthy.
http://www.earthmagazine.org/article/hazardous-living-maps-according-geologists




#if y0ure wondering what all the comments about beer are about #remember that geologists are just slightly taller than average dwarves #with flannel

 

So I just got back from a talk on how granites form.

Apparently, Earth is pretty unusual in the solar system for having granites/high silica, low Mg/Fe rocks. Venus is all basalts and relatives, and Mars has some high-silica stuff, but it apparently looks like it only formed during the very earliest parts of the planet's history and there's not much.

All that is really interesting when compared to Earth's bimodal crust -- that is, we have both basaltic, dense, oceanic crust which forms basins (that are then filled by oceans, natch) and granitic/felsic continental crust which is a lot lighter and floats up, giving us, yanno, actual land. My understanding, although don't quote me on this 100%, is that Earth has more than enough liquid water to cover the entire surface of the planet kilometers deep with if we had just a little less topographic relief.

On top of that, we know that the mantle's chemistry is pretty damn mafic -- that is, if you melt it and erupt it, it's going to cool as basalt.
So where the hell are all these continent-forming light granites (and not-quite-granite friends) coming from?

The classic model relies on the fact that the dominant minerals in basalt (olivine, amphiboles, pyroxenes and some plagioclase feldspars) crystallize at higher temperatures than the dominant minerals in granite (quartz and alkali feldspars). The idea is that you get a big chamber of molten basalt-forming magma and just let it sit under the surface of the earth and let it cool. As it cools, it starts crystallizing your olivines and high-temperature plagioclases, and those crystals settle to the bottom of the big liquid magma pool. Eventually, so the story goes, those high-temperature minerals, which tend to be high in Mg and Fe, pull all the Mg and Fe out of the liquid and leave you with a bunch of molten silica plus some random junk, ready to cool into a granite (or get erupted on the surface as a rhyolite).

theeeere's a couple problems with this, though. First off, it requires that these cooling crystals are actually able to sink and thus remove themselves from the magma in question. If you look at the fluid dynamics, this works just fine in relatively runny basalt magmas, but as you get into higher-silica magmas, the fluid just gets too damn viscous for crystals to sink pretty much at all.

Secondly, you notice how much I emphasized "liquid" up there? Well, we can "see" fluid under the Earth's surface. Earthquakes produce, among other kinds of waves, shearing body waves that will travel through solids but not fluids. This is how we know that the outer core of the Earth is liquid and that the mantle is not, by the way, seismic stations more than 104 degrees of arc away from the epicenter of an earthquake won't pick up s-waves from it, because they are in the shadow of the outer core. If there were bigass blobs of fluid magma, like the above model requires, sitting in the middle of the crust, we would be able to detect their (much smaller) seismic shadows. The biggest we've found is a ~100m thick lens of molten magma sitting right under a (basalt erupting) mid-ocean ridge. Nooot quite the granite making slow cooker we're looking for.

Also, for petrological reasons that I am sorry to say I couldn't really follow, we can tell that this sort of "settle out the high Mg/Fe crystals" thing hasn't happened in situ for any of the big blocks of granite we can go look at, which would require that it happen somewhere else in the lower crust, then all that nice ready-to-make-granite magma would have to be transported elsewhere to actually make the granite, leaving all that Mg/Fe rich mush behind, which apparently isn't consistent with out observations of crust density (I think. I didn't follow this part too well. haven't had petrology yet)

Anyways, this is apparently all a cause for some shenanigans and a bunch of competing hypotheses in the granite fandom, especially since this has been the working model for I think over half a century.

The speaker I was listening to was proposing a model of differentiation of melts along a temperature gradient. He had run some experiments with a rock-level pressure cooker thing: you stick a buncha rock powder of the desired chemistry into a capsule, and put it in a piston thing to crank up the pressure to the correct amount for the geological process you're looking at, then run a current through it to melt it. The way he'd set up the experiments was a gradient in the capsule from 950C on one end to 350 C on the other, and dumped in a bunch of powdered andesite (intermediate between granite and basalt) and about 4% water by weight (relatively typical for this kind of magma, I think. maybe a little wet, but definitely something you could find in the wild). And then he left it for a couple months.

What he wound up with was a gradient from Mg/Fe rich liquid at the hot end of the capsule (that is, it quenched to amorphous glass when the capsule was cooled, indicating that it was still liquid at the point the machine was switched off), then through a magnetite-rich semi-molten layer, then a amphibole and pyroxene rich semi-molten gabbro mush, then progressively more silica-rich and less molten to the bottom which was 95% crystallized granite. Note that under the classical model of granite formation, you don't have any liquid under 650 degrees C, period.

He also did a super clever thing to analyze the remaining molten 5% of that coolest side without loss of data when the ting was switched off and cooled. He heat-shocked a bunch of quartz grains and loaded them into a mixture which was saturated with respect to quartz (so the grains wouldn't just dissolve and homogenize with the rest of the melt). Heating the quartz means that it can heal the cracks from heat shocking it, but as it does so, some of the melt gets into the cracks and gets trapped, so it can by analyzed later.

He found that the melt in the granite side of things was 20-50% water (!), so like borderline between "magma" and "hydrothermal fluid." Also cool thing, each of the trapped bits of melt had a bubble of the same size, that shrunk and disappeared when it was heated back to 350 C, meaning that the melt was homogeneous at 350 C and that it didn't start to degass until it had cooled further.

He also found some interesting things with isotope ratios, with the cooler side of the capsule being more enriched in the heavier isotopes of Si, Fe and Mg (despite being lighter overall, due to lower overall Fe and Mg), which matches up with what has been observed in the field in terms of isotope ratios from granites versus related gabbros and basalts.

Given all this, plus some petrological stuff I didn't understand, plus some physics models of atoms in melt, he's hypothesizing that the differentiation of rock types across this temperature gradient is caused by water-aided diffusion at relatively low temperatures, on a scale of meters to tens of meters.

By this model, then, a bigass granite plutolith wouldn't be cooling from a glob of molten magma all at once, it would be a relatively thin sill or dike of molten magma that came in and cooled, then another one came in below it, so there's a temperature gradient from cool on top to hot below, and thus migration of atoms by diffusion on a meter to tens of meters scale, basically replicating what he found in his capsule. Then more magma coming in layer by layer continues to move that temperature gradient like in the experimental capsule down, leaving behind a bunch of granite. He has a field site in the Andes that seems to support this model, too, and that matches up with global isotope trends, suggesting that it is fairly typical.

So that's one ball in the court of "where the fuck did this granite stuff come from," anyways. (it doesn't work for all situations, though. the speaker guy mentioned that for something like yellowstone you do need a huge source of liquid magma, although I don't know that we see anything like that under yellowstone at this particular moment).

 

lemme know if i need to clarify anything there, by the way.
It was getting long, and the sentences were already p messy, so I left in more jargon than I would have liked to otherwise.

 

That's fascinating, and I had no idea that this was still something uncertain.

 

I love how so much of Science is:

(Insert Question about anything, really)

Science: I dunno! I honestly have no idea beyond 3 theories, 5 conflicting dissertations from 9 different countries and 12 thousand milligrams of caffeine let's figure it out.

 

Oh! I forgot to mention:


this guy has also tried to do similar temperature gradient experiments using ground up basalt as the starter rock (for a more basaltic melt).
They all blow up in a couple weeks. Predictably enough that he was able to guess when to within a day.

Keep in mind that this is super durable equipment, and should totally be able to handle a sudden increase in pressure from off-gassing or whatnot, and all these eruptions are blowing out of this tiny little hole in the capsule where the wiring to melt the rock connects.

If you add more than 4 weight percent water, they blow up within days.

He has absolutely no fucking idea why.

 

Small question for space peeps

SO you know how there are symbols used to represent each planet.
Are there symbols to represent the planet types?
Like, is there something you could slap onto a picture of mars that means "this is a terrestrial planet"
Just like you could take a really zoomed in picture of jupiter, and not everyone might know that it's jupiter, so you slap Jupiter's symbol on it and they go "ohhh It's jupiter, not saturn)

 

I wouldn't think so. The symbols come from alchemy and astrology, which were mostly developed long before we figured out terrestrial planet versus gas giant (around the 1950s). So unless astrologers suddenly introduced a bunch of new symbols in the last 60 years (and they might have, I don't know), I don't think that distinction would have been made.

 

Okay, so there arent official ones, so since I needed some, I made some. They arent the best, but they are for one small town school project.

They wont be on this card, these are just my examples (also the project is in french so yeah writing is in french)
Terrestrial, gas giant, ice giant

Thoughts?

 

@Imoyram maybe the terrestial and gas giant ones shouldn't both be brown? yeah, they are different browns and coloured in a different way, but i can see it being confusing for people, especially from a distance.

 

Hmm. Distance might not be an issue, because these are going to be about half the size they are rn, so about 1cm x 1cm.

I see what you're getting at though