Printable Version of Topic

Posted by: Paolo Apr 6 2013, 10:10 AM

a new Explorer satellite dedicated to exoplanets around near stars

http://www.nasa.gov/home/hqnews/2013/apr/HQ_13-088_Astro_Explorer_Mission_.html

Posted by: JRehling Apr 10 2013, 06:52 PM

TESS will look for exoplanets around nearby stars. The catch with the transiting method is that geometry doesn't do us any favors: Any planets which don't transit can't be seen, and the farther the planet is from the star, the less likely a favorable alignment is.

The probability of a transit varies from over 20% for planets with periods of a few days to less than 1% for planets at a distance of about 1 AU. So even if 100% of stars have a planet at 1 AU, you'd have to monitor hundreds of stars to see a few such planets. And when you talk about hundreds of stars, you're no longer talking about "nearby."

Kepler addresses this by looking at many stars farther away -- Kepler stars are basically between about 500 and 6500 light years away.

TESS will monitor about two million stars looking for planets with periods of less than two months. Once such systems are identified, that'll provide many promising leads for follow-on studies with other instruments.

Posted by: AndyG Apr 10 2013, 11:15 PM

You are, relatively: there are around 500 stars up to the 40 light year mark. That's practically next door.

Andy

Posted by: vjkane Apr 11 2013, 02:54 AM

Anyone know how TESS will identify transits since it will be scanning across the sky and most transits will occur outside the immediate field of view?

Posted by: JRehling Apr 11 2013, 03:09 AM

Some of TESS's targets will be closer, but the typical star observed will be a couple of hundred light years away.

It's an interesting question as to what the closest transiting planet is. Luck will figure large in that. There's no reason in principle why a very close one couldn't be, but the random orbital inclinations will bear whatever fruit they do.

Posted by: alphasam Apr 11 2013, 06:28 AM

If I remember correctly It will scan the same three fields in turn every 30 minutes (i.e ten minutes integration each) for a couple of months, giving it a slightly shorter measurement cadence than Kepler, and then it will move on to the next set of three fields and repeat.

Sadly this "Step and Stare" method means it will only be sensitive to short-period planets, ones close to their stars. It may be possible to detect temperate worlds around M-dwarfs though.

Posted by: JRehling Apr 15 2013, 10:17 PM

While more knowledge is always (?) better, there are losses in completeness due to geometry and signal-to-noise even when monitoring is allowed to run longer.

A major contribution will be to identify which systems have inner planets that are aligned favorably. Those will merit follow-on observations to look for the rarer outer planets that are aligned favorably. It's useful as triage so a later mission might focus on a smaller number of systems.

Also, note that outer planets may still show single transits even during a very short period of observation, and so given large enough numbers, many outer planets will appear as candidates in the data, with more loosely-bound parameters.

Posted by: JRehling Aug 10 2015, 08:12 PM

As TESS gets closer to launch, more complete information about this exciting mission is online. A quick overview is here:

http://tess.gsfc.nasa.gov/science.html

In essence, TESS will observe stars that are, typically, about ten times closer and 100 times brighter in apparent magnitude than the stars that Kepler observed. In a two-year primary mission, it will observe almost every bright star in the sky for 27 days. It will observe roughly 10% of them for about 81 days, and about 3% of them for a year.

Compared to Kepler, it will observe more stars for a short period, but fewer stars for a long period (Kepler observed most of its targets for 3 years), and much closer stars, on average. Because it searches the whole sky, it will also possibly find systemic differences in, e.g., galactic latitude, that Kepler, aimed at one tiny portion of the sky, missed.

The geometry of TESS's search is based on the ecliptic, with a gap along the ecliptic, and the stars that it observes for the longest centered at the ecliptic poles, in Draco and Dorado. These stars, conveniently, may also be observed at any time by JWST, which is also restricted from viewing near the ecliptic because of bright things like the Sun and the Earth in the way.

TESS should provide an almost comprehensive survey of "hot" planets with favorable transiting geometry in the solar neighborhood. This is interesting in that it will identify a large number of systems that have long-period transiting planets in the same orbital plane as the hot planet, but with too long a period for TESS to identify.

It will also find many warm planets and probe the habitable zone of cooler stars that happen to be located relatively near the ecliptic pole. Many of its discoveries will be promising targets for follow-up science by the JWST.

It is not particularly well-suited to finding planets with ~1-year periods orbiting G stars like the Sun. An extended mission of two years may provide that possibility for stars near the ecliptic pole, but I'm not sure which options would be possible for the extended mission. Focusing on one celestial hemisphere (north or south) may open up more discoveries with a long period while neglecting the other hemisphere. Or, giving equal attention to both hemispheres may allow medium-period discoveries everywhere.

Posted by: JRehling Aug 12 2015, 05:39 PM

The other night, I saw Sara Seager of MIT speak about exoplanet discovery programs, which spanned the work of Kepler in the recent past, to TESS in the near future, and the Starshade mission that she's working on for later on.

One takeaway that I found interesting was the idea that the search for "earthlike" planets, as she sees it, will be strongly focused on planets of Earth's size and larger orbiting red dwarfs. And this is for the inescapable factor that they are easier to find. Not only do detection methods favor such planets, as opposed to planets orbiting G stars with a period of about a year, but they are more numerous objectively speaking. In fact, virtually any survey designed to find planets orbiting G stars with a period of about a year would end up finding a much larger number of such planets orbiting red dwarfs much closer in anyway.

A planet orbiting, say, Proxima Centauri, would receive earthlike levels of energy from its star if it had a period of 6 days. Thus, the TESS survey, looking at all sufficiently bright red dwarfs for 27 days, will see all the planets that transit such stars in their HZs, given low enough noise (which is assured for larger planets).

This goes a long way to explain the importance of TESS. There's a excellent chance that whenever we are able to spectrally examine an earth-sized (or small super-earth-sized) planet in its star's HZ, that TESS will be the mission that will have discovered that planet in the first place so that the follow-up could take place.

Posted by: alphasam Aug 13 2015, 07:19 PM

The problem is "Earth-like" has a HUGE number caveats here. Many would argue Earth-size worlds in the so-called "habitable zone" around M dwarfs have little chance of being that Earth-like. This is the problem of stretching the "Goldilocks zone" concept too far, which is solely designed to discuss temperatures.

If you take a look at some M dwarf lightcurves from Kepler you will immediately see one problem. They have a very distinctive characteristic appearance, which is caused by their extreme magnetic activity. They have flares and coronal mass ejections that make the Sun look as calm as a millpond. These events spew huge quantities of radiation out into space, and because planets in the Goldilocks zone of M dwarfs have to be so close, they are right in the firing line.

Secondly, a planet that close to its star is almost certainly going to be tidally locked, producing extremes of temperature and consequently horrifying convectional winds. Thats if it even has an atmosphere left after being battered by the raging stellar wind. It's very likely to become a barren wasteland very quickly.

So yeah these planets are going to be heavily studied because of their abundance and ease of discovery, but probably best not call them Earth-like. NASA indeed is focusing heavily of them, but for example ESA with Cheops and Plato are going for more benign K and G dwarf planets, this an area of vigorous debate right now.

Posted by: JRehling Aug 14 2015, 06:08 PM

All completely correct, alphasam. Except as a convenient shorthand, the term "earthlike" has a number of noxious aspects, whereas its only two merits are its brevity and the fact that it sells newspapers. For the time being, the only three qualities we can ascertain about a transiting planet of this kind are:

1) Its size.
2) Its level of radiation energy received.
3) The environment its star provides (e.g., circular or elliptical orbit; likely tidal lock; flares vs. none).

Mass is a possible property, although for planets of Earth's size, this is difficult to impossible to measure in most cases.

Ideally, we would find all the exoplanets within 50-100 lightyears or so that are "earthlike" in all those ways, but even to find all such planets that transit their stars would require a massive survey. Roughly speaking, we'd need to monitor all such stars continuously for 4 years. This is about 25 times as much observing as TESS' primary mission will perform, so it won't be cheap, and it won't happen soon.

For each "earthlike" planet according to (1) and (2) we find around a sunlike star, we're likely to find about 50 around red dwarfs, so we're going to get to know the red dwarf cases in useful numbers long before we get to know the ones around sunlike stars. Getting a survey of what the red dwarf "earthlike" (1) and (2) planets' spectra look like will be a profoundly exciting result; not as exciting as the sunlike star "earthlike" planet survey, but surely sooner in coming.

Posted by: Hungry4info Aug 14 2015, 08:52 PM

A minor caveat about (1), which I'm certain you already know but worth posting for public consumption. It's more precise to say that we know the star-planet radius ratio, as opposed to the actual planet size. The precision of the planet radius is dependent on the precision of the stellar radius (and how accurate your limb darkening coefficients are, but that's typically a small effect). For some systems, that can contribute to the planetary radius being perhaps several tenths off, in others the radius can be constrained exquisitely.

Posted by: hendric Aug 17 2015, 05:26 PM

Found an interesting website on an observation program "Living with a Red Dwarf":

http://www.astronomy.villanova.edu/lward/index-old.htm

http://arxiv.org/pdf/1111.2872v2.pdf is a better link to their paper than from the website itself.

Posted by: Holder of the Two Leashes Jul 27 2018, 08:27 PM

TESS has begun science operations.

https://phys.org/news/2018-07-nasa-tess-spacecraft-science.html

https://spaceflightnow.com/2018/07/27/tesss-planet-hunt-begins/

Posted by: Holder of the Two Leashes Dec 20 2018, 08:10 PM

Some papers relating to TESS discoveries so far:

https://arxiv.org/abs/1809.05967

https://arxiv.org/abs/1809.07242

https://arxiv.org/pdf/1810.02341.pdf around a very young F/G type star that is part of a binary star system.

The first public release of data is already available https://archive.stsci.edu/prepds/tess-data-alerts/. It mentions over forty good candidate for exoplanets already detected.

Posted by: JRehling Jan 4 2019, 06:22 AM

There's a lot to pore over here! Just perusing casually, I see case 259962054 looks like a really interesting terrestrial planet candidate. As I see the numbers, it has an equilibrium temperate a bit cooler than Mars and a radius of 1.2 Earths – certainly within the realm of possibility of earthlike conditions.


Many months of reading Kepler data tables makes this a lot more explicit to me than it otherwise would be. They are presenting the data in Kepler-like formats, which is great, because that framework was very well thought out.

It's going to be great to get more of this data, a somewhat more condensed summary of the information, and more evaluation of the candidates.

Posted by: JRehling Jan 5 2019, 04:27 PM

Having taken a bit more time to review the data release, with a mind towards locating the more earthlike candidates, I'll emphasize the following:

There are about 40 stars that have a planet candidate with a period of 8 or more days. I focused on these since shorter periods are very likely to be very hot, even for small M dwarfs.

TESS, for those who don't know, will look at most of the entire sky, but will look, in the primary mission, at most of the sky for only about one month, and can therefore only find, in those parts of the sky, planets with quite short periods. However, the observed segments overlap, and in the overlap, planets with longer periods can be found. This data release covers three observational periods, so the longest possible period offering three transits is about two months.

In this data, I found five planet candidates worthy of being called somewhat earthlike, including two in the same system. Of those, none really hits the Goldilocks area of parameter space "just right." We have some too big, some too hot, and some too cold, but given our lack of understanding of how planets evolve, I think these five are worth of mention. I'll list simply the star's TESS ID, magnitude, orbital period (days), planet radius (in Earths), and equilibrium temperature (K).

12421862 9.85 20.4 1.6 325
37749396 8.43 13.5 1.6 466
12421862 11.24 3.8 1.1 379
12421862 11.24 10.6 1.4 268
259962054 12.17 52.0 1.2 171

Important reference: The equilibrium temperatures of Venus, Earth, and Mars are 301, 255, and 207, respectfully. Only the fourth of those planets falls in that range. However, remember that these parameters have considerable uncertainty, and we have no good knowledge of what a climate might be for any of them.

The variable I'd most like to add, and can perhaps derive, is distance in light years, because these are all obviously quite close compared to any of the Kepler discoveries. Note that "magnitude" isn't a simple constant for each star, but depends upon the wavelengths to which a telescope responds.

To put this further in context, the TESS main mission will have 26 such observational periods, of which this data includes only 3, so we may naively expect a 9x increase in the results, but that both understates and overstates the matter in various ways. For the stars that TESS observational periods overlap, we will also explore outward, seeing longer periods, and that may turn up additional earthlike planets at greater distances from hotter stars. However, some candidates may prove not to be real.

Still, this gives us confidence that tens of generally earthlike planets will be found by TESS, and these are particularly exciting in terms of the potential for follow-up science. These, along with the handful of terrestrial planets already know to orbit nearby stars, will be the systems that JWST and the ELT will observe and maybe give us actual observed data for the planets' atmospheric composition and temperature. This is the start of a very interesting decade, and some of those planets on the list above may become very famous in the next few years.

Posted by: antipode Jan 5 2019, 09:17 PM

Very interesting taster of things tp come!

12421862 11.24 10.6 1.4 268 looks especially interesting.

Are we likely to see discovery papers soon on the arxiv, or will they wait til RV followup?

P

Posted by: Hungry4info Jan 5 2019, 09:28 PM

Depends on the system. All of these candidates can be found https://archive.stsci.edu/prepds/tess-data-alerts. This site is regularly (assuming the U.S. government is not shut down) updated as TESS data rolls in. Others will work to confirm them with RV.

GJ 143 + HD 23472, https://arxiv.org/abs/1812.04501
Pi Men, https://arxiv.org/abs/1809.05967
LHS 3844, https://arxiv.org/abs/1809.07242
HD 2685, https://arxiv.org/abs/1811.05518
HD 202772, https://arxiv.org/abs/1810.02341
HD 1397, https://arxiv.org/abs/1811.01882
HD 219666, https://arxiv.org/abs/1812.05881

Altogether nine planets have been confirmed so far (two at HD 23472).

Which correspond to GJ 7, GJ 1008, LHS 1140 (TIC 92226327) and 2MASS J02520450-6741155. Both planets at LHS 1140 have been discovered (and confirmed) prior to TESS.
LHS 1140 b https://arxiv.org/abs/1704.05556
LHS 1140 c https://arxiv.org/abs/1808.00485

Posted by: JRehling Jan 6 2019, 06:46 PM

That's a fantastic update (backdate!). How did you unify the TESS results with the existing discoveries? RA and Dec are given for each star, so you could do it that way. The number of nearby red dwarfs is not limitless, but I'd like to have a method to merge the data automatically if possible, as TESS data pours in.

The LHS 1140 system is about 40 light years away, the same as TRAPPIST-1, which, with ten planets between the two systems, will make ~40 light years an interesting distance threshold in discussions of nearby exoplanets.

Posted by: Hungry4info Jan 6 2019, 07:01 PM

Just going through SIMBAD and searching each RA+Dec. If you want I can e-mail you an Excel document containing each TOI and their respective ID where known.

Posted by: Hungry4info Jan 8 2019, 01:56 AM

And now TOI-197.01 (HIP 116158 / HD 221416) is confirmed.
https://arxiv.org/abs/1901.01643

Posted by: Hungry4info Jan 29 2019, 02:01 AM

Five low-mass planet candidates orbiting TYC 8856-192-1 (TOI-125), two of which have been confirmed.
https://arxiv.org/abs/1901.09092

Posted by: Hungry4info Jan 30 2019, 11:20 AM

An Eccentric Massive Jupiter Orbiting a Sub-Giant on a 9.5 Day Period Discovered in the Transiting Exoplanet Survey Satellite Full Frame Images
https://arxiv.org/abs/1901.09950

TOI-172 = TYC 6932-301-1

Posted by: JRehling Aug 17 2020, 02:16 AM

As of July 2020, TESS has completed its primary mission and is now beginning an extended mission that will approximately repeat the main mission.

While the observations of the primary mission are complete, the analysis is not nearly complete. Having found only 66 confirmed new planet discoveries so far, this will inevitably increase considerably even if no new observations occurred, although I couldn't begin to estimate what the right multiplier is for the future yield.

However, the extended mission provides the opportunity for completely new kinds of discoveries. TESS only viewed most of its primary mission targets for 27.4 days each, which means that a transiting planet with an orbital period less than 9.2 days was guaranteed to exhibit three transits during the observations, a planet with an orbital period of 9.2 to 13.7 days might exhibit three transits, and a planet with a longer orbital period could not exhibit three transits, but a planet with an orbital period up to 27.4 days might have exhibited two transits.

It's the planets in that range from 9.2 to 27.4 days that will benefit from the extended mission, when more observation time makes it very likely that planets with one or two previously recorded transits will be observed in a third or second and third transits. So, the window of orbital periods around each star will widen considerably. This will moreover mean that we can find candidate earthlike planets orbiting larger and brighter stars with the extended mission data. Even better, a second extension would suddenly widen that enormously, creating opportunities for the discovery of candidate earthlike planets orbiting sunlike stars. In essence, the primary mission lets us find candidate earthlike planets circling M dwarfs, the extended mission will push that out to brighter, hotter M dwarfs, and a second extended mission would push it out to K and G dwarfs.

So, if TESS has finished a third of its eventual observations, we may have only a minuscule fraction of the interesting discoveries in hand now. This is quite different from a mission like a planetary orbiter where one expects much of the scientific value to come with first observations.

By the way, TESS's observational windows overlap in some weird but fortuitous ways, so some stars have already been observed enough to allow discoveries in longer orbital periods. But the best is almost certainly yet to come.

Posted by: antipode Aug 17 2020, 04:23 AM

I almost get the feeling that the discovery rate thus far is a little disappointing? Is that so? Does it accord at all which what was predicted?

I'm sure there are many many more discoveries to be pulled out of the data, even without an extended mission, which will probably morph into an extended extended mission anyway ($$$?)

P

Posted by: Hungry4info Aug 17 2020, 10:39 AM

TESS has identified over 2,000 candidate planets. The problem is that ground-based RV follow-up is resource-intensive. It has always been the bottleneck of the process toward getting these kinds of planets confirmed.
https://exofop.ipac.caltech.edu/tess/

Posted by: JRehling Aug 17 2020, 04:46 PM

I don't think the results are disappointing. The results are just at a very early stage of processing, despite the fact that the observations of the main mission have concluded. A few weeks before Kepler's main mission concluded, analyses of its data had found 105 confirmed planets. That number ended up at 2,662. (Coincidentally, the candidate planets as of early 2013 was about the same as the confirmed planets as of now.)

Thanks to Kepler, we know the intrinsic frequencies of various star-planet combinations. There's no reason to suspect that the TESS sample will differ significantly from what was expected based on that. All we have to hope for is that operations are successful, which has been the case so far.

Luck will apply when expected findings for a particular type of planet is in the single digits, and then we may get lucky and find, say, six, or unlucky and find one. But it will be a few years from now before we know how that goes.

Posted by: antipode Aug 18 2020, 10:11 AM

Youre right - I keep forgetting that candidates need to be confirmed by ground based RV, and that's a slow process.

P