Super-Dense Planets... or perhaps just rather large

[major_clanger]

io9 has an intriguing story about a planet larger than Earth and as dense as lead. (Not to mention being white hot).

A bit of digging turns up the paper that this story is based on, made available last week. However, in the course of searching for it I found this paper by another team, which whilst referencing the first paper (I assume they saw a draft) uses IR rather than optical measurements to come up with a diameter for 55 Cancri e that is half as much again as that derived in the other paper. That cuts the density dramatically for the same mass (which is fairly well-characterised) thus taking the planet from 'twice as dense as Mercury' to 'about as dense as Mars', which is a little less dramatic.

What both papers make clear, and which the io9 story doesn't, is that even if 55 Cancri e is as dense as lead, it isn't made of lead. Or comparably-dense silver or thorium, or even, alas, iron with a core of platinum. Rather, it's made of iron, but iron compressed to much greater than normal density by the sheer mass of the planet, some eight times that of Earth. Incidentally, a planet that close to its star couldn't be made of lead - at over 2,000 degrees, it would have boiled away...

So, be a little wary of dramatic science claims. It may be that someone has missed another paper, or just picked the one with the more dramatic conclusions.

Comments

[steer.livejournal.com]

To be fair to your original source io9 it is pretty clear they're claiming it's the density of lead not made of lead and it is an accurate report of the first paper.

Certainly no fair on calling them for missing the second paper you cite which was published on arxiv at 19:25pm the day before they published their story -- even if they monitor arxiv daily with RSS feed readers (as I do) they'd have probably not found out about it until after their story was out.

Neither paper is published yet so I wouldn't take bets on which was true just yet. (The estimate for diameter from the second isn't "half as much again" but 1.3 times as much).

The first paper has the surface temperature estimate as below the boiling point of lead, the second paper has it just above (but both assume an albedo of zero). In reality though if the planet were largely composed of lead (which it obviously is not) it would, I guess, head to the centre (as with iron in the earth).

[major-clanger.livejournal.com]

Fair call on the timing - it was bad luck for io9 that a second paper should come out soon with a somewhat less dramatic set of results. However, I'd still say that the io9 article, while an accurate summary of the first paper, does not make it very clear that whilst the planet (on its analysis) is as dense as lead, it is very unlikely to be made of lead; from the comments it's evident that many readers assume that this is in fact what the planet is implied to be made of.

As for the contrasting results, it may of course be that they bracket the correct value, in which case 55 Cancri e is somewhat (but not dramatically) denser than Earth or Mercury, which would be expected for a dense terrestrial planet of 8 Earth masses close enough to its star for all its volatiles and many of its lighter minerals to have boiled off.

[steer.livejournal.com]

As for the contrasting results, it may of course be that they bracket the correct value

They don't I'm afraid... admittedly I don't recall the meaning they assign to the +/- though 95% confidence interval would be my assumption, in which case I guess that it is not so amazing to find no overlap even if none of them have made an error in their assumptions or modelling.

does not make it very clear that whilst the planet (on its analysis) is as dense as lead, it is very unlikely to be made of lead

I guess you're right since people in the comments are mistaken. I thought it was clear (since "density of" and "comparable to" are mentioned) but then I've some basic knowledge of relative cosmic abundance and know the unlikelihood of a planet actually made of lead.