It's reasoned guess. They know three things about the planet: its mass, its size (radius/volume), and its orbital distance. They also know some of the chemistry of the star it orbits.
From the mass and radius, they inferred the density of the planet is a lower than other rock planets. They can explain low density in several ways. One is a thick layer of low-density supercritical water on the surface. Very large, extremely hot "oceans". An alternative is low-density graphite or diamond inside the mantle.
What they show is that a carbon interior is plausible given the data. They model several things. For one, the planet's formation from a primordial dust disk -- given the star's chemistry, which is unusually carbon-rich, and physical models of how elements separate out in accreting to planets. And they model the planet's current interior, based on known physics. They can't directly observe the planet chemistry; they can only model it.
Where are the measurements from?
The planet's radius/size was measured photometrically when it passed between us and its star (an occlusion, like an eclipse). They don't have the resolution to see the planet, but they can measure the reduction in the star's brightness when it occludes, and hence how big of a cross-section it has.
The mass was discovered by measuring the gravitational effect on its star. It is a very tiny effect -- it's too small to see any difference in the star's position. But small changes in velocity shift the frequency of its emitted light, and that can be detected. This is a very small effect: the star's velocity variation due to this planet is just 6.3 meters/second, and it was measured with 0.2 meter/second precision! That's how they know the mass to within 5% accuracy.
The star's element composition is known by spectrometry -- measuring how much of an atom or ion is at its surface, by how much that atom absorbs light of a specific color.
You can do it yourself. Do you know why the sky is blue? Because blue light waves are scattered by molecules in our atmosphere. The light emanating from distant stars also interact with the molecules in the star and nearby planets etc.
We see the wobble in the star via the frequency of it's light changing.
My understanding is that the assumption of diamond is made by making many measurements of the absorption spectra of the star to try and get an idea of what kind of matter is floating around it as dust and gas and from that trying to work out what the balance of chemistry in that system is likely to be.
You can also look at the star's spectra during a transit of the planet and use that to work out information about the atmosphere and surface.
At the end of the day, they do not know what is inside the planet, this is just their best guess based on what they can measure.
Presumably they inferred its mass from its orbital motion, and its size from how much light it blocks off as it crosses in front of its star. Density is smply mass divided by volume. Then they compare it to the density of known materials and find that only graphite and diamond are a good match.
http://arxiv.org/pdf/1210.2720.pdf
It's reasoned guess. They know three things about the planet: its mass, its size (radius/volume), and its orbital distance. They also know some of the chemistry of the star it orbits.
From the mass and radius, they inferred the density of the planet is a lower than other rock planets. They can explain low density in several ways. One is a thick layer of low-density supercritical water on the surface. Very large, extremely hot "oceans". An alternative is low-density graphite or diamond inside the mantle.
What they show is that a carbon interior is plausible given the data. They model several things. For one, the planet's formation from a primordial dust disk -- given the star's chemistry, which is unusually carbon-rich, and physical models of how elements separate out in accreting to planets. And they model the planet's current interior, based on known physics. They can't directly observe the planet chemistry; they can only model it.
Where are the measurements from?
The planet's radius/size was measured photometrically when it passed between us and its star (an occlusion, like an eclipse). They don't have the resolution to see the planet, but they can measure the reduction in the star's brightness when it occludes, and hence how big of a cross-section it has.
http://en.wikipedia.org/wiki/Transiting_extrasolar_planets#T...
The mass was discovered by measuring the gravitational effect on its star. It is a very tiny effect -- it's too small to see any difference in the star's position. But small changes in velocity shift the frequency of its emitted light, and that can be detected. This is a very small effect: the star's velocity variation due to this planet is just 6.3 meters/second, and it was measured with 0.2 meter/second precision! That's how they know the mass to within 5% accuracy.
http://en.wikipedia.org/wiki/Radial_velocity_method
http://arxiv.org/pdf/1208.5709v1.pdf
The star's element composition is known by spectrometry -- measuring how much of an atom or ion is at its surface, by how much that atom absorbs light of a specific color.
http://spiff.rit.edu/classes/phys230/lectures/spec_interp/sp...