32 new planets discovered, some of them gas giants

[Akkleptos]

Please excuse if this has come up before... A quick scan of the recent topics didn't reveal anything related to this relatively new discovery. Mods and Admins feel free to relocate or delete, should it be the case.

European scientists find trawl of 32 new planets

WASHINGTON (Reuters) - European astronomers announced they had found 32 new planets orbiting stars outside our solar system and said on Monday they believe their find means that 40 percent or more of Sun-like stars have such planets.

The planets range in size from about five times the size of Earth to about five times the size of Jupiter, they said. More have been discovered, too, they said, promising more announcements later this year.

The latest discoveries bring the total of known exoplanets to about 400, said Stephane Udry of the Geneva Observatory in Switzerland.

"Nature doesn't like a vacuum so if there is space to put a planet it will put a planet there," Udry told reporters in an Internet briefing from a meeting of astronomers in Porto, Portugal.

"More than 40 percent of stars like the sun have low mass planets," Udry added.

The team used the High Accuracy Radial Velocity Planet Searcher or HARPS, a spectrograph attached to the European Southern Observatory's 3.6-metre (11.8-foot) telescope in La Silla, Chile.

The spectrograph does not image the planets directly but scientists can calculate their size and mass by detecting tiny changes in a star's wobbling caused by a planet's small gravitational pull.

Astronomers are keen to find Earthlike planets as these are the most likely to harbor life. HARPS has spotted 75 planets circling 30 different stars. The ESO team did not give details of which stars the 32 new planets were circling.

(Reporting by Maggie Fox, editing by Jackie Frank)

Does this mean that planets are far more common than we previously suspected? Also, what is the size-limit on a gas giant, before its own mass makes it collapse and maybe start stellar fusion? And, what I guess everybody's asking... does this discovery make the probability of more Earth-like planets out there more feasible?

[Junghalli]

Also, what is the size-limit on a gas giant, before its own mass makes it collapse and maybe start stellar fusion?

Minimum mass for hydrogen fusion is thought to be 75-80 Jupiter masses (reference).

Personally I'm impatient for results from Kepler.

[cosmicalstorm]

I remember this article from early summer, these two pieces of news in combination would lead me to believe that at least primitive life might be very common.

Life’s First Spark Re-Created in the Laboratory

A fundamental but elusive step in the early evolution of life on Earth has been replicated in a laboratory.

Researchers synthesized the basic ingredients of RNA, a molecule from which the simplest self-replicating structures are made. Until now, they couldn’t explain how these ingredients might have formed.

“It’s like molecular choreography, where the molecules choreograph their own behavior,” said organic chemist John Sutherland of the University of Manchester, co-author of a study in Nature Wednesday.

RNA is now found in living cells, where it carries information between genes and protein-manufacturing cellular components. Scientists think RNA existed early in Earth’s history, providing a necessary intermediate platform between pre-biotic chemicals and DNA, its double-stranded, more-stable descendant.

However, though researchers have been able to show how RNA’s component molecules, called ribonucleotides, could assemble into RNA, their many attempts to synthesize these ribonucleotides have failed. No matter how they combined the ingredients — a sugar, a phosphate, and one of four different nitrogenous molecules, or nucleobases — ribonucleotides just wouldn’t form.

Sutherland’s team took a different approach in what Harvard molecular biologist Jack Szostak called a “synthetic tour de force” in an accompanying commentary in Nature.

“By changing the way we mix the ingredients together, we managed to make ribonucleotides,” said Sutherland. “The chemistry works very effectively from simple precursors, and the conditions required are not distinct from what one might imagine took place on the early Earth.”

Like other would-be nucleotide synthesizers, Sutherland’s team included phosphate in their mix, but rather than adding it to sugars and nucleobases, they started with an array of even simpler molecules that were probably also in Earth’s primordial ooze.

They mixed the molecules in water, heated the solution, then allowed it to evaporate, leaving behind a residue of hybrid, half-sugar, half-nucleobase molecules. To this residue they again added water, heated it, allowed it evaporate, and then irradiated it.

At each stage of the cycle, the resulting molecules were more complex. At the final stage, Sutherland’s team added phosphate. “Remarkably, it transformed into the ribonucleotide!” said Sutherland.

According to Sutherland, these laboratory conditions resembled those of the life-originating “warm little pond” hypothesized by Charles Darwin if the pond “evaporated, got heated, and then it rained and the sun shone.”

Such conditions are plausible, and Szostak imagined the ongoing cycle of evaporation, heating and condensation providing “a kind of organic snow which could accumulate as a reservoir of material ready for the next step in RNA synthesis.”

Intriguingly, the precursor molecules used by Sutherland’s team have been identified in interstellar dust clouds and on meteorites.

“Ribonucleotides are simply an expression of the fundamental principles of organic chemistry,” said Sutherland. “They’re doing it unwittingly. The instructions for them to do it are inherent in the structure of the precursor materials. And if they can self-assemble so easily, perhaps they shouldn’t be viewed as complicated.”

[MKSheppard]

So Gas Giants are very common? That leads me to believe that exinction level events are not very common -- because one of the roles that Jupiter and the other supermassive giants play is being a gravity sink and target sponge for asteroids to suck them in before they can hit Earth.

[Dave]

So Gas Giants are very common? That leads me to believe that exinction level events are not very common -- because one of the roles that Jupiter and the other supermassive giants play is being a gravity sink and target sponge for asteroids to suck them in before they can hit Earth.

Selection bias. Gas giants are easy to detect, basically because they're so big.

[Simon_Jester]

Regardless of selection bias, if we see gas giants orbiting a large fraction of randomly selected stars, gas giants remain very common. Selection bias means that we can't say anything about the frequency of Earthlike planets (which we can't see), but it doesn't affect our ability to measure how common the planets we can actually see are.

[ThomasP]

Funny enough, this link came in on my RSS feed right as I started reading this thread.

The basic ingredients for life have been found around a second extrasolar planet, scientists reported Tuesday.

Although the planet itself is not habitable by life as we know it, the discovery could mean that the basic components of life are widespread in the atmospheres of many kinds of exoplanets.

The new find was made by training both the Hubble and Spitzer Space Telescopes on HD 209458b, a hot Jupiter that orbits very close to its sun-like star. It’s located 150 light years away in the Pegasus constellation. In December of last year, Swain’s team found a similar Jupiter-like planet, HD 189733b, with carbon dioxide in its atmosphere.

“Detecting organic compounds in two exoplanets now raises the possibility that it will become commonplace to find planets with molecules that may be tied to life,” team leader and Jet Propulsion Laboratory astronomer Mark Swain said in a press release.

The study of exoplanets has exploded since the first were discovered in the early 1990s. Just Monday, astronomers announced the discovery of 32 new exoplanets. And detections aren’t just growing in number, but sophistication as well. Exoplanetary scientists are learning more and more about the systems in which the planets are found.

Early exoplanet discoveries were made using a variety of techniques, but primarily by measuring the “wobble” a star exhibits in the presence of another massive body. In more recent years, scientists have looked for “transiting” planets, which pass in front of and behind their stars. Far more can be learned about these celestial bodies.

When an exoplanet passes in front of its star, scientists are able to translate small differences in the color of the light arriving at Earth into a chemical signature for the planet’s atmosphere. For example, HD 209458b has water and carbon dioxide, just like HD 189733b, but it’s also got a lot more methane.

“The high methane abundance is telling us something,” said Swain. “It could mean there was something special about the formation of this planet.”

Planetary spectroscopy is easiest to do for systems in which a large exoplanet orbits very close to its home star. With smaller planets orbiting farther from their star, it is to detect the minute changes in the star’s light.

Though the Kepler Space Telescope is likely to find many Earth-like planets, it could be a decade before we have the technological capability to definitively detect a rocky planet with an atmosphere and orbit like ours, an Earth twin.

edited to fix quote tags

[Mayabird]

So Gas Giants are very common? That leads me to believe that exinction level events are not very common -- because one of the roles that Jupiter and the other supermassive giants play is being a gravity sink and target sponge for asteroids to suck them in before they can hit Earth.

Probably common, but most of the ones we've found are much, much closer to their stars than Jupiter. But as previously mentioned, it could also be selection bias - it's much easier to see the wobble when it's visible over a period of days or weeks than when the orbit's measured in years or decades.

[Simon_Jester]

True. What I'd very much like to know is the actual number of stars we've examined for evidence of planets.

If we discover 32 Jovian planets in close orbit around 32 out of 10000 stars, it says little about the likelihood of planets existing around a random star, or what kind of planets will exist. If we discover 32 Jovian planets in close orbit around 32 out of 100 randomly selected stars, it says quite a bit.

[Mayabird]

True. What I'd very much like to know is the actual number of stars we've examined for evidence of planets.

If we discover 32 Jovian planets in close orbit around 32 out of 10000 stars, it says little about the likelihood of planets existing around a random star, or what kind of planets will exist. If we discover 32 Jovian planets in close orbit around 32 out of 100 randomly selected stars, it says quite a bit.

I'm pretty sure it's just "have the computer compare the nightly images taken from this slice of sky and sort out the ones that show the wobble." Digital photography has made it much, much easier since they can get images of much dimmer stars that wouldn't affect the photography plates but it hasn't been available for very long so there's not as much data to fall back on for the long-orbit gas giants. The first definite exoplanets weren't even discovered until the 1990s - limits of technology and not as many people looking for them. And again, it's much easier to see a definite periodic wobble over the course of days than to wait several years to make absolutely sure it's a wobble for a longer-orbit planet, instead of just a weird flaring event or something.

Everyone knows what I'm talking about when I say "wobble" here right? Should I just explain it so nobody feels embarrassed?

[Elaro]

The image produced by the light of a star and an orbiting planet would seem to wobble? Did I understand correctly?

[GrandMasterTerwynn]

The image produced by the light of a star and an orbiting planet would seem to wobble? Did I understand correctly?

She probably should've just explained it. It's simple physics. An orbiting planet exerts the exact same amount of gravitational force upon its parent star that the parent star exerts upon the planet. This means that the planet and star orbit about a common center of mass. Since a star is orders of magnitude more massive than a planet, the center-of-mass of the system resides very close to (but not precisely at) the center of the star. So while the planet executes a large orbit around this common center of mass, the star completes its own very tiny orbit, thus, seeming to wobble; completing periodic oscillations back and forth across the trajectory it would normally take if there weren't a planet tugging on it.

Since this wobble is very tiny, we typically detect it it by measuring the Doppler shift of the star's light as this wobble causes the star to move slightly towards, and slightly away from us, inducing a slight blue- and redshift of the star's spectrum. To ensure that you've got a "real" reading, you have to take measurements over many days . . . enough to cover several of the putative planet's orbits around it's parent star (so we can establish that there is an oscillation, and determine its period. This tells us that there's a planet in orbit around the star, and how long it takes to orbit. We can also determine the planet's mass, by measuring how much Doppler shift the planet is causing . . . since the heavier the planet is, the further out the common center-of-mass is, and the bigger the resulting gravitational wobble.)

This, incidentally, is why we find lots and lots of large planets orbiting really close to their parent stars. They're big enough to induce a Doppler shift measurable by our instruments, and orbit quickly enough that scientists can confirm that there's a planet orbiting the star.

[starslayer]

I'm pretty sure it's just "have the computer compare the nightly images taken from this slice of sky and sort out the ones that show the wobble." Digital photography has made it much, much easier since they can get images of much dimmer stars that wouldn't affect the photography plates but it hasn't been available for very long so there's not as much data to fall back on for the long-orbit gas giants. The first definite exoplanets weren't even discovered until the 1990s - limits of technology and not as many people looking for them. And again, it's much easier to see a definite periodic wobble over the course of days than to wait several years to make absolutely sure it's a wobble for a longer-orbit planet, instead of just a weird flaring event or something.

Since my present work involves this, I'll give some insight here:

The way we usually do observing runs nowadays goes something like this:
1. Generate a list of stars that we are interested in, for whatever reason. Usually these reasons are a) they're not blue (or large and red), and b) they're close by.
2. Establish how often we want to observe a certain star on our list; this is based on several factors which I have not learned, but you end up with what we call a "cadence." If a star has a cadence of X, we want to observe it every X days, telescope time and weather permitting.
3. Sit in a remote ops room for the entire night from sunset to sunrise, and call in our next target via video feed to the telescope operations room. As soon as the exposure is done, we get the data on our computers. It used to be only a few years ago that most astronomers had to travel to observatories; not anymore.

The reason we generate and use those lists is because we cannot take spectra of a section of sky without hopeless amounts of background noise - we must take a spectrum of a single star at a time. We initially take the light cone from the telescope's main optics and pass it through an iodine cell (a glass tube with some gaseous iodine in it) to provide a spectral reference on the exposure, and then pass it through the various slits and diffraction gratings. Without this reference, we cannot establish a Doppler shift. Once we have our data, we follow the process that Terwynn laid out.

As for the instrument itself, our spectrometers (such as HIRES, if you want an example of a specific instrument) have gotten better and better over the years, gaining more and more resolution. Actual light throughput has never really been much of an issue, even with faint M dwarfs. Today we can detect a stellar wobble of only a few m/s. This does not allow us to detect Earth II; however, the next generation of detectors, which will be online in the next couple of years, will be capable of detecting motions of mere tenths of a m/s, more than sensitive enough to see an Earth-like planet in an Earth-like orbit. It will still likely be a decade or two before such a planet is finally announced, because even assuming we start observing one tomorrow, as Terwynn mentioned, we need many complete periods to confirm a detection.

As a historical note, Geoff Marcy and his team, the first to actually hunt for planets, got good data on 51 Pegasi in the late 80's, IIRC. They just didn't realize they'd found the first exoplanet before the Swiss team released their results in (IIRC) '95. They all indulged in a big collective facepalm after that.

[Simon_Jester]

Starslayer, could you offer some insight into how many stars are searched to find a given number of exoplanets of the typical size? As I mentioned above, that's a fairly important number that is not often released in the second and thirdhand accounts of exoplanet searches seen in the general media.

Everyone knows what I'm talking about when I say "wobble" here right? Should I just explain it so nobody feels embarrassed?

I, for one, understand perfectly. For that matter, there are multiple types of wobbling that you could (theoretically) look for, so using the general term "wobble" probably makes sense. While you normally use radial wobble and detect it via Doppler shift, as I understand it you could in principle use the transverse back-and-forth wobble. It's not as good, reliable, or easy as using Doppler shift, though.

[starslayer]

Starslayer, could you offer some insight into how many stars are searched to find a given number of exoplanets of the typical size? As I mentioned above, that's a fairly important number that is not often released in the second and thirdhand accounts of exoplanet searches seen in the general media.

Unfortunately, I'm mostly a grunt; right now I'm looking through old data for some troubling anomalies we've been seeing (we suspect they're instrumentational, rather than having anything to do with the stars in particular). My (very rough) estimate for how many stars we search through to find a gas giant is probably less than a hundred. Keep in mind that this only includes stars already likely to have planets; we do not observe blue stars and red giants/supergiants, nor the larger and more massive dwarfs (the Sun is a yellow dwarf, so I am talking relatively hefty stars here). Basically, first and foremost we discard stars around which life could not possibly exist, and stars which never live long enough for planets to form. Other surveys may look at any star or SNR for planets - for instance, one of the first multiple planet systems was discovered around a pulsar.

I, for one, understand perfectly. For that matter, there are multiple types of wobbling that you could (theoretically) look for, so using the general term "wobble" probably makes sense. While you normally use radial wobble and detect it via Doppler shift, as I understand it you could in principle use the transverse back-and-forth wobble. It's not as good, reliable, or easy as using Doppler shift, though.

If I understand your meaning here, by transverse wobble you mean perpendicular to our line of sight, yes? If so, no we cannot detect that. Such a wobble does not show up in a spectrum, and astrometry instruments are not nearly sensitive enough outside of spectra to detect such a small change in position (for a hot Jupiter, the motion would be a hundred thousand kilometers or so over an entire orbit), and likely never will be.

[Simon_Jester]

All right; I have a tolerably good sense of the scale of wobbling involved, but very little sense of the limit of modern astrometric resolution. I'm not surprised to be wrong about that.

[And_Atom_JT]

I'm afraid this site hasn't been updated in a while, but it's still a good guide to all of the exoplanets discovered up to 2005, and has some nice artwork to boot.

http://www.extrasolar.net/evmain.asp

If a link has already been posted I apologize, but I thought it would be a nice thing to contribute.

[Surlethe]

The other thing you can do is measure the decrease in flux caused by a transit. This only works for systems where the Earth is in the orbital plane, but the same holds (generally) for measuring with a Doppler shift (i >> 0 leads to an unmeasurable Doppler shift, as starslayer outlined). We can get satisfactory measurements of that from the ground - our observatory has measured 51 Pegasi's transit, for instance - but it's much better from space: Kepler early results.

Edit: It looks like the sinusoidal variation in the star's light curve is a result of the heating of the gas giant's atmosphere. How cool is that? We can measure an exoplanet's temperature.

Edit2: To GMT's point above, here is a distribution of discovered planets. Note that they are either very massive (1e2-1e4 earth masses) or very close to the star (<1 AU orbital radius). That's the major selection bias at work.

[starslayer]

It looks like the sinusoidal variation in the star's light curve is a result of the heating of the gas giant's atmosphere. How cool is that? We can measure an exoplanet's temperature.

Yes, it is; for a planet with no atmosphere, you would expect a trapezoidal light curve for the eclipse (you can't really see that in the figure, because planets are so small compared to a star). By comparing the radius of the planet to that we would expect from polytropic models (which are more than good enough for this), and taking its absorption spectra to find out what the make-up of the atmosphere is, we can then determine its temperature down to the point where its atmosphere is opaque (optical depth of 2/3). Even if such spectra are not available, we can still make an approximate temperature measurement.

By taking the absorption spectra of an exoplanet, we can also immediately determine whether it has Earth-like life, by looking for free oxygen and methane in significant quantities. Seeing water vapor would seal the deal.

[Surlethe]

And, of course, we ought to note that O2 would not only indicate likeness to Earth, it would indicate life: the only reason we know why a planet would have significant molecular oxygen in its atmosphere is because life is metabolizing it from things like CO2.

[Modax]

I'm curious; how big would a visible light telescope have to be to directly image an earthlike planet (and see a disk rather than a speck) from many lightyears away?

Suppose you could convert the Death Star 2's parabolic dish into a giant reflector telescope. That would give a mirror of ~200km diameter. What could be seen with such a device?

[GrandMasterTerwynn]

I'm curious; how big would a visible light telescope have to be to directly image an earthlike planet (and see a disk rather than a speck) from many lightyears away?

Suppose you could convert the Death Star 2's parabolic dish into a giant reflector telescope. That would give a mirror of ~200km diameter. What could be seen with such a device?

If we had a magical mirror made from unobtanium and blessed by the Elven gods, floating out in free space, the resolving limit of our 200 km telescope can be given by angular resolution in radians = wavelength / diameter (in meters) of telescope objective. Plugging the numbers in gives us a theoretical angular resolution of 2.9x10-12 radians. This assumes that we only care about yellow light, which has a wavelength of 580 nm.

We can then (as a simple approximation) solve for the smallest resolvable diameter at a given distance via the Pythagorean formula. An earth-sized planet can be directly resolved to a disc at a distance of . . . one light-year, and as a point source out to 100,000 lightyears with a long enough exposure and integration time.

So, the short answer is: Unrealistically huge. To directly image Earth-sized planets, you would need what is called an interferometer; which is a precisely placed collection of small telescopes. The resolving power of an interferometer is dictated by the length of its baseline, which is (simply put) the diameter or length of the interferometer array. However, it's not going to be an image like you'd be thinking of since: A) An interformeter has the resolving power of an equivalent gigantic telescope, but a miniscule fraction of the light-gathering capability. B) Since an interferometer only gathers light along the points in its array; a linear interferometer will produce a set of diffraction fringes. (An actual telescope produces them too. If you look at a star through a telescope, you will notice that it's got a tiny bright spot corresponding with the diffraction maxima, and a set of diffraction rings surrounding this maxima where the photons constructively interfere, separated by gaps where they cancel each other out.)

[Junghalli]

The main site has an optical lens calculator which is designed to answer exactly this question.