Eight exoplanets close to star

Gwynston

BBC article: Star system has record eight exoplanets

Another nice Kepler discovery, but the transit method seems to mostly pick up planets that are very close to their parent star.

I guess this is the nature of the method? If I understand it correctly, it works by analysing dips we see in the light of a star as its orbiting planets pass in front.

So we're kind of cherry picking so far all the examples that have obvious, frequent dips we can spot quickly. By their nature, they are likely to be stars with close, frequent orbiters.

So up to now we've not spotted a lot of systems with farther out, less frequently orbiting planets, purely because we're not likely to see them so often?

And of course, we can only see these transits in systems where the plane of the orbits just so happens to be lined up with our line of observation. I guess the vast majority of systems probably operate in plane which means we simply can't observe transits from our viewpoint?

ps200306

Gwynston wrote: »

BBC article: Star system has record eight exoplanets

Another nice Kepler discovery, but the transit method seems to mostly pick up planets that are very close to their parent star.

I guess this is the nature of the method? If I understand it correctly, it works by analysing dips we see in the light of a star as its orbiting planets pass in front.

So we're kind of cherry picking so far all the examples that have obvious, frequent dips we can spot quickly. By their nature, they are likely to be stars with close, frequent orbiters.

So up to now we've not spotted a lot of systems with farther out, less frequently orbiting planets, purely because we're not likely to see them so often?

And of course, we can only see these transits in systems where the plane of the orbits just so happens to be lined up with our line of observation. I guess the vast majority of systems probably operate in plane which means we simply can't observe transits from our viewpoint?

Yes, you are right on all counts. The probability of a planetary transit occurring in a system randomly oriented with respect to our line sight is approximately the radius of the star divided by the radius of the planetary orbit. So, for example, the Earth's orbit is roughly two hundred times the radius of the Sun. The probability of far away aliens seeing the Earth transiting is therefore about half a percent (1/200).

Then the difficulty of spotting the transit is related to the depth of the dip in the light curve. The dip is proportional to the ratio of the area of the planet's disc to that of the star, and therefore to the square of the ratio of the planet's and star's radius. Jupiter is a tenth of the radius of the Sun, so a transit of Jupiter would cause a 1% dip in the light curve (1/10 x 1/10). The Earth is ten times smaller again, so an Earth transit would cause a 0.01% dip. A larger orbital radius also increases the number of times you have to image a star to be likely to catch a planet in the act of transiting.

The reason why the Kepler search is so fruitful when the probabilities are so low is that it can image a large number of stars simultaneously, about 150,000 of them, and it does this over and over again for a period of years. Also, because planets in a system tend to have roughly coplanar orbits, detecting one planet in a system somewhat increases the chances that others will be found. The sensitivity of the Kepler photometry is about three times as good as is needed to detect an Earth-sized planet orbiting a Sun-sized star.

When multiple transit observations of a planet are made they can be "phase-folded" (i.e. combined on top of each other) to improve the signal to noise ratio. A sharp light curve allows for things like grazing transits to be used to glean even more information about the star and planet. Also, variations in the timing of successive transits can be used to infer the existence of other planets in the system from their gravitational influences. And finally, the reflected light alone from a close-in Jupiter sized planet is enough to cause a signal just about within Kepler's sensitivity limits, and this has recently been used to find even non-transiting planets in the Kepler data.

murphthesmurf

It's amazing that we have found so many planets with the odds so small for everything lining up for us to see the dip in light. Yet from my Google search we have found 2771. The other method of looking for a stars wobble hasn't been as fruitfull with 658 found. It is a shame the wobble hasn't turned up more, as it would presumably be our best chance to find planets which are not in our line of site of the star as in the shadow method.
To think a few years back it was presumed that stars with planets were rare. I'd say stars without planets are rare.

Hector Savage

ps200306 wrote: »

Yes, you are right on all counts. The probability of a planetary transit occurring in a system randomly oriented with respect to our line sight is approximately the radius of the star divided by the radius of the planetary orbit. So, for example, the Earth's orbit is roughly two hundred times the radius of the Sun. The probability of far away aliens seeing the Earth transiting is therefore about half a percent (1/200).

Then the difficulty of spotting the transit is related to the depth of the dip in the light curve. The dip is proportional to the ratio of the area of the planet's disc to that of the star, and therefore to the square of the ratio of the planet's and star's radius. Jupiter is a tenth of the radius of the Sun, so a transit of Jupiter would cause a 1% dip in the light curve (1/10 x 1/10). The Earth is ten times smaller again, so an Earth transit would cause a 0.01% dip. A larger orbital radius also increases the number of times you have to image a star to be likely to catch a planet in the act of transiting.

The reason why the Kepler search is so fruitful when the probabilities are so low is that it can image a large number of stars simultaneously, about 150,000 of them, and it does this over and over again for a period of years. Also, because planets in a system tend to have roughly coplanar orbits, detecting one planet in a system somewhat increases the chances that others will be found. The sensitivity of the Kepler photometry is about three times as good as is needed to detect an Earth-sized planet orbiting a Sun-sized star.

When multiple transit observations of a planet are made they can be "phase-folded" (i.e. combined on top of each other) to improve the signal to noise ratio. A sharp light curve allows for things like grazing transits to be used to glean even more information about the star and planet. Also, variations in the timing of successive transits can be used to infer the existence of other planets in the system from their gravitational influences. And finally, the reflected light alone from a close-in Jupiter sized planet is enough to cause a signal just about within Kepler's sensitivity limits, and this has recently been used to find even non-transiting planets in the Kepler data.

Brilliant post!
So much better than any of the articles I've read on the mainstream news sites!

ps200306

murphthesmurf wrote: »

It's amazing that we have found so many planets with the odds so small for everything lining up for us to see the dip in light. Yet from my Google search we have found 2771. The other method of looking for a stars wobble hasn't been as fruitfull with 658 found. It is a shame the wobble hasn't turned up more, as it would presumably be our best chance to find planets which are not in our line of site of the star as in the shadow method.

Yep, the wobble method (a.k.a. radial velocity method) is exquisitely sensitive. It looks for Doppler shifts in the emission lines in the star's spectrum. But because it relies on spectroscopy you can generally only do one star at a time, whereas Kepler can do tens of thousands by the photometric method. The radial velocity method is fruitful if you already know what you're looking for, so it's used on all the Kepler candidates, and a Kepler transit is not considered confirmed until it has been followed up with RV.

The great thing about the RV method is that the size of the wobble is proportional to the mass of the planet. The downside is that spectroscopy only tells us the component of the wobble in our direction (hence the "radial" in RV). The fact that we can't see the planet directly and thus don't in general know the orientation of its orbit (and which exact direction it is pulling) means that the mass is only constrained to within a factor of the sine of the angle of inclination of the orbit. But when we can combine RV and transit methods we get the best of all worlds:

• The RV method gives us the mass exactly, as sine i ≅ 1 by definition for a transiting planet.
• The transit dip gives us the radius.
• The combination of the two gives us the density.

So we know if we are looking at a gas giant or a rocky planet.

murphthesmurf wrote: »

To think a few years back it was presumed that stars with planets were rare. I'd say stars without planets are rare.

Yes, it's totally mind-blowing. We might have suspected that planets were common due to theories about how stars form from collapsing gas clouds, but we had no way of confirming this until Kepler. All of a sudden we know there are probably hundreds of billions of planets just in our own galaxy.

Gwynston

ps200306 wrote: »

And finally, the reflected light alone from a close-in Jupiter sized planet is enough to cause a signal just about within Kepler's sensitivity limits, and this has recently been used to find even non-transiting planets in the Kepler data.

Wow - so we can actually even directly observe some exoplanets? :eek:
Have there been some examples of that already?

ps200306

Gwynston wrote: »

Wow - so we can actually even directly observe some exoplanets? :eek:
Have there been some examples of that already?

I don't think this method would be considered direct observation, as it still relies on changes in light levels. Direct observations generally imply that the object is optically resolved, that is, you can form a distinct image separate from the parent star. That's coming too, though!

To answer your question -- a recent paper "discovered" some dozens of additional non-transiting planets from the Kepler data. They used some clever machine-learning algorithms to sift out the extremely noisy signals. They will still need radial velocity measurements to confirm all the candidates but the initial signs are that the effort has been pretty successful (> 90% detections are real).

In terms of actual direct observation, that is a much trickier proposition as planets are so incredibly dim compared to their host stars. Also, at the huge distances to stars, the maximum angle between star and planet is tiny, typically a handful of milliarcseconds even for nearby systems. But scientists have been working on figuring out what they're going to be able to see with the next generation of giant telescopes, such as the 39 metre E-ELT. A number of known planets in nearby systems should be resolvable. The neat thing about this is that we can then use spectroscopy to study the constituents of their atmospheres, among other things.

murphthesmurf

In a documentary a few years back debating the possibility of life on other planets one scientist came up with his worst case scenario as follows. (I forget who the scientist was)
'If 1% of the known stars have planets orbiting them, 1% of those are in the habitable zone, and 1% of those actually harbour life. Then that is over 30,000 planets with life'. This documentary was aired possibly 10 yrs ago. Given what we are learning now 30,000 may not scratch the surface for our own Galaxy alone.
Exciting times we live in.

ps200306

murphthesmurf wrote: »

In a documentary a few years back debating the possibility of life on other planets one scientist came up with his worst case scenario as follows. (I forget who the scientist was)
'If 1% of the known stars have planets orbiting them, 1% of those are in the habitable zone, and 1% of those actually harbour life. Then that is over 30,000 planets with life'. This documentary was aired possibly 10 yrs ago. Given what we are learning now 30,000 may not scratch the surface for our own Galaxy alone.
Exciting times we live in.

I would guess the scientist was Frank Drake. He came up with the famous Drake equation, which combines a whole bunch of different factors for the probability of intelligent life. We've now filled in the blank for the likely number of planetary systems. We're hazier on the size of the habitable zone because of things like most stars being red dwarfs, which produce a lot of nasty (potentially sterilising) radiation. And there are many other factors that we still don't have much of a clue about. But we're closer than we were!

Hector Savage

One of my favourite posts about the SETI project

https://waitbutwhy.com/2014/05/fermi-paradox.html

Gwynston

Ben Miller's book on the subject is a good, accessible read and covers Drake, Fermi and many other considerations, like how life actually gets started:

The Aliens Are Coming!: The Exciting and Extraordinary Science Behind Our Search for Life in the Universe

murphthesmurf

Hector Savage wrote: »

One of my favourite posts about the SETI project

https://waitbutwhy.com/2014/05/fermi-paradox.html

Mind blowing stuff. As to the theories about why we haven't found intelligent life, I'd kind of suspect they're all correct to some extent. A combination of all the theories.
To bend the mind even further, did any of you see the Horizon documentary last night on BBC Four? It looked into the, what some consider, increasing possibility of a 'multiverse' and some of the possible ways in which it could exist. So not only is there the possibility of other life in our vast universe, but also in an infinite number of other universe.

ps200306

Hector Savage wrote: »

One of my favourite posts about the SETI project

https://waitbutwhy.com/2014/05/fermi-paradox.html

Yeah, that one's a great piece of writing, and I really like this conclusion: "whatever the truth actually is, it’s mindblowing".

ps200306

Gwynston wrote: »

BBC article: Star system has record eight exoplanets

I've just been chasing up the published paper on which the BBC article is based. It's in Astronomical Journal, with a preprint available on arxiv here.

Very interesting! I mentioned that machine learning had been used to detect large non-transiting planets in tight orbits by their reflected light. But of course, the same techniques can be used to detect smaller, further out, planets that do transit. You're still just looking for statistical inferences about very faint signals in very noisy data. That's how this 8th planet discovery around Kepler-90 was done, using machine learning developed by Google.

Hector Savage

murphthesmurf wrote: »

Mind blowing stuff. As to the theories about why we haven't found intelligent life, I'd kind of suspect they're all correct to some extent. A combination of all the theories.
To bend the mind even further, did any of you see the Horizon documentary last night on BBC Four? It looked into the, what some consider, increasing possibility of a 'multiverse' and some of the possible ways in which it could exist. So not only is there the possibility of other life in our vast universe, but also in an infinite number of other universe.

I love the idea of Van Nuymen probes (sp) ... Thats why I agree with Brian Cox on this - for sure there is other life in the universe (maybe even in our own solar system) but we could be the only intelligent life.