I am favoring iron because:
- Because you find it in the oddest places, where it shouldnt be in stellar models. See "Mysterious Iron Factories in the Early Universe."
- Because that's what certain myths say.
- Because of the running difference images.
- Because of the coronal loops.
- Because of its magnetostrictive properties.
- Mass fractionation data. Here is Oliver Manuel's Iron Sun page. I dont really agree with his idea of the interior of the sun being a neutron star, but I find his mass fractionation data interesting. http://www.thesunisiron.com/
Other than that it's a hollow sphere with the right dimensions to receive the 160 minute (?) cycle wave length would be my first guess.
The depth of the surface was based on opacity and the location of White Light flares.
In the standard model, the solar moss extends into the chromosphere. If you ask me the solar moss is just a part of the solid surface below the photosphere that is at a different potential and is therefore glowing.
I'm convinced that there is a lot more iron, near the surface, than my model has. The huge volume of iron in CMEs, and its behavior, just doesn't look like the 1 iron per every 30,000 hydrogen atoms that the quiet sun spectroscopy registers.
My model has a way of collecting heavier elements around active regions, and if a flare occurs, we'll see a misrepresentative abundance of iron in the CME. But I still think that my model comes up shy. As Michael notes, 171 Å imagery (such as
this) shows such fine detail that 1 part in 30,000 just isn't going to get it.
Robitaille is saying that the supercritical hydrogen is forming into a graphite-like liquid crystal, and that this prevents heavier elements from settling. I'm not sure about the liquid crystal, but I'll easily concede that mass separation will be greatly impeded if there is a supercritical substrate. This means that heavier elements that somehow get to the surface might stay there much longer than mass separation would have otherwise allowed. So kamikaze comets and/or nuclear fusion in solar flares might put heavier elements there, and then they just accumulate.
From the ballistics of coronal rain, we easily observe that iron really doesn't like being expelled. I think that there is a powerful electric field there, and iron's higher degrees of ionization make it particularly susceptible to that field. So an H I atom or an He I/II atom might get accelerated out into the solar wind, but an Fe X atom gets pulled forcefully back to the Sun.
But since the granules appear to be well-mixed, and since the quiet sun spectroscopy sees a lot of hydrogen and helium, and not a lot of iron, I'm thinking that the bulk of the iron starts at 4800 km below the tops of the granules. Though I gave Michael a bunch of shit about the SDO First Light imagery, I actually think that he's probably right. If so, it agrees with my qualitative statements about the hydrodynamics of the granules (i.e., that the thermal bubbles come from a depth that is 4~5x deeper than they are wide, meaning a depth of 4000~5000 km), and Kosovichev's helioseismology. I just don't know how to prove it.