≡ Menu

TESS: A Full-Sky Exoplanet Survey

The news that NASA has approved the TESS mission kept my mood elevated all weekend. TESS (Transiting Exoplanet Survey Satellite) has been the logical NASA follow-up to Kepler ever since the Space Interferometry Mission was canceled in 2010. The point is that Kepler looks at a field of stars with the goal of developing a statistical analysis, helping us (ultimately) to home in on the value for ηEarth (Eta_Earth), the fraction of stars orbited by planets like the Earth.

To do this, Kepler is looking out along the Orion Arm of the galaxy, with almost all the stars in its field of view between 600 and 3000 light years away. In fact, fewer than one percent of Kepler’s 156,000 stars are closer than 600 light years. There are plenty of stars beyond 3000 light years, but as we push beyond this distance, the stars become too faint for Kepler’s transit methods to be effective. The carefully chosen field in Cygnus and Lyra is ideal for Kepler’s statistical data but the next question to ask is how many Earth-like planets are to be found around relatively nearby stars.


TESS is led by principal investigator George Ricker (MIT Kavli Institute for Astrophysics and Space Research), whose team will be working with an array of wide-field cameras to perform an all-sky survey, as opposed to the intense ‘stare’ Kepler makes at a particular starfield. Ricker says the mission should be able to identify thousands of new planets in the solar neighborhood in a survey that will cover 400 times as much sky as any previous exoplanet mission.

We all have ideal missions we’d like to see fly, from the Space Interferometry Mission to some sort of grandly designed Terrestrial Planet Finder, but among the realistic options in front of us, TESS makes abundant sense. It will home in on small rocky planets and help us measure their mass, density, size and orbit, along with offering up data on their atmospheres. And you can also think of TESS as something of a pointer scope for the James Webb Space Telescope and other future instruments that will begin the hunt for astrobiological signatures on other worlds.

Image: TESS’s primary goal would be to identify terrestrial planets orbiting nearby stars. Credit: MIT Kavli Institute for Astrophysics & Space Research.

“The TESS legacy,” says Ricker, “will be a catalog of the nearest and brightest main-sequence stars hosting transiting exoplanets, which will forever be the most favorable targets for detailed investigations.” That makes the $200 million funding for this bird money well spent. In an interview on the Kavli Foundation site, Ricker adds that the interactions of multiple planets as they orbit their host star, spotted through transit timing variations, can aid the search:

Measuring transit time variations is particularly important because if you find a large Neptune- or Saturn-sized planet that appears to be tugged a little bit, you can make an estimate of how massive the planet is that is actually doing the tugging. Measuring transit time variations, which is a technique enabled by Kepler data, is a bit like successively unnesting a Russian matryoshka doll – you go a few layers down and you can find smaller and smaller planets in the system. That’s one of the things that doing transit time variations, during the TESS mission, will enable us to do.

In other words, we’ll have TESS spotting planets around stars bright enough for ground-based telescopes to home in on these transit time variations (TTV), something we can’t readily do with most of the Kepler planets because their stars are too faint. The system is designed to detect changes in the intensity of a star’s light down to 40 parts per million. By comparison, Earth viewed from outside the Solar System would cause a transit drop of about 85 parts per million. Ricker estimates that TESS will be able to detect as many as 2700 planets, including several hundred Earth-size worlds. The TESS mission is now scheduled for launch in 2017.


Comments on this entry are closed.

  • RonSmith April 8, 2013, 11:17

    I don’t understand one aspect of the TESS mission. It is designed to only detect planets with a period of 2 months or less. I really don’t see what the point of that is, such planets would be searingly hot if orbiting a K class star, and tidally locked if orbitting an M class star. In neither case would they seem like good candidtates to me for follow up study. Or am I misunderstanding something?

  • Robert Feyerharm April 8, 2013, 13:06

    This is fantastic news – but why can’t NASA engineers change the orientation of the Kepler Telescope to point at new targets?

  • Paul Gilster April 8, 2013, 13:07

    Ron, this is what Dr. Ricker says in a different part of the interview I quoted:

    What we envision is that after two years we will complete a census of all the transiting extrasolar planets in the sky that have a period of less than about a month. This will be particularly valuable for studying the small M stars that are so abundant in our galaxy. We will have a good sample of planet periods extending out to about a year, and this will cover a wide variety of solar types – from stars that are hotter than the sun to stars that are a lot colder than the sun, and for which the habitable zones range from small distances from the stars to large distances from the stars.

  • RonSmith April 8, 2013, 13:24

    Yes, I read that quote, but I still don’t get it. Perhaps I’m being dense today.
    “We will have a good sample of planet periods extending out to about a year” doesn’t make sense to me. I don’t know the details of the mission plan for TESS, but the fact that it is only intended to conduct a survey for 2 years, and is intended to be a ‘full sky’ survey, means it will see very few repeated transits, and then only of planets of a period of less than 2 months, apparently. How can they claim to end up with any sort of reliable, confirmable, data for planets with longer periods? Mind you, I am not trying to say that they don’t know what they are talking about, rather I am complaining that media coverage of TESS seems to be lacking some essential information for the story to make sense, though I do not know what that information is.

  • Daniel April 8, 2013, 13:50

    Far as I know,unfortunately TESS won’t be design to detect transit planets in M dwarf Large Sample,because M dwarf stars Isn’t bright enough,then many of this Stars with a potential of short period Earth-size planets in HZ will be missing.that would be good Follow up JWST

    I read in somewhere that TESS will search only 1000 of M dwarf at 30 pc of distance,this is ever less that the Kepler M dwarf sample than is about ~3800

    In my opinion I think that this is not good enough and mission should work on the Near IR like the ELEKTRA Mission (that few people Know about,and I don’t know why this mission option haven’t show up by the media)


    This is like a “TESS in Infrared”,that would be much more efficient in detect Earth-size planets in the HZ of the most numerous nearby stars and most common stars in universe the Red dwarfs,and that mission would cover brown dwarf and all the other stars types search for planet transit that TESS will cover.

  • Thomas Mazanec April 8, 2013, 14:05

    Actually systems with transiting exoplanets will not “forever be the most favorable targets for detailed investigations.” since eventually we will have solar gravity lens focus missions and/or hypertelescopes which can survey planets in any system.

  • andy April 8, 2013, 14:05

    Well habitability is certainly one major area of interest, but there’s no reason it should be the only factor in exoplanet studies. Consider e.g. SuperWASP and HATNet which are basically hot Jupiter machines, and still going strong.

    Furthermore M-dwarf planets are excellent choices for follow-up: firstly the planet is larger relative to the star, secondly and perhaps more importantly it would be really nice to have some actual observations about whether such planets can support habitable conditions, rather than just relying on theoretical considerations going back-and-forth on the matter.

  • Daniel April 8, 2013, 14:50

    I just hope that NEAT mission get approval

    NEAT: a space born astrometric mission for the detection and characterization of nearby habitable planetary systems”


    and the Starshade for JWST or other telescope design for it ,with this 2 missions get approval, I think will fully greatly cover all our Nearby Exoplanets in HZ or not,around all types of stars at a least 10 pcs

    Then with E-ELT that can take image of Earth-size planets in our Nearby M dwarf HZ (and high precision radial velocity of 1 cm/s with CODEX and other giants ground based telescope Like TMT )


    and with EUCLID mission that already get approval to be launch in 2019 that should detect planets by microlensing

    ExELS: an exoplanet legacy science proposal for the ESA Euclid mission I. Cold exoplanets



    and the GAIA Mission of course,and now TESS that is very important to detect our nearby transit planets (ever thought that in Near-IR would be much better)

    In my opinion ever After the mission that have been cancel like SIM and TPF mission, next decade,could be a gold decade for exoplanetary science alter all.

    If approve NEAT and the Starshade will be great (this 2 missions would Cover the gap of SIM and TPF missions left )

  • coolstar April 8, 2013, 15:19

    TESS makes so much sense that I’m a little surprised (but very pleased) that NASA selected it. And it’s price, it’s obviously a steal. I don’t hold a lot of hope for JWST greatly advancing our knowledge of exo-planets though. It’ll do (if it a) ever flies and b) works as designed, very big IFS) some atmospheric studies of a very small number of exo-planets, and not much else in the field. The extra-galactic bullies aren’t going to let people interested in stars and planets get very much time on it, you can be sure of that! This would be different if NASA would design and launch a Starshade to work in conjunction with JWST, but that’s not going to happen until the space based interferometry people stop killing those proposals, and THAT isn’t going to happen until those folks retire or die off. I expect a LOT more for the next generation of giant ground based systems, with spectrographs with better than 10 cm/sec RV precision and ultra-high resolution adaptive optics.

  • tom April 8, 2013, 17:40

    The only thing that bothers me about these missions is the timelines. Putting the launch date past the next presidential election, how quick can this be mothballed for a new administration funding cut? Potentially, there is plenty of room for a better atlas for our galaxy… but projects like these really need to be moving faster to get the new expertise to the youngest professionals; that’s the actual value of doing the science.

  • Paul Gilster April 8, 2013, 19:32

    Ron S. writes:

    How can they claim to end up with any sort of reliable, confirmable, data for planets with longer periods?

    I’m also trying to dig into this — like you, I’m a bit puzzled by what I’ve been able to find so far re planets with longer orbital periods.

  • Enzo April 8, 2013, 19:56

    Also, in 2017, CHEOPS from ESA :

    It seems to concentrate on studying known transiting exoplanets around bright nearby stars with a small telescope. I wonder if there is some overlap here and the usual lack of coordination between the two agencies.

  • tesh April 8, 2013, 21:23

    Paul G, Ron S

    How can they claim to end up with any sort of reliable, confirmable, data for planets with longer periods?

    I’m also trying to dig into this — like you, I’m a bit puzzled by what I’ve been able to find so far re planets with longer orbital periods.

    My take on this is that during the 2 month window, they will either see a transit or not. For Sun like stars, they may see no orbits at all or up to 12. They will definitely see no repeat transits of any of the planets. Most likely they will see a transit or two and from that flag it for further studies. Also, from what I have seen at PH.org, one transit is enough to get a rough idea of period and radius, so very quickly they can get a “picture” of the system architecture. For stars smaller than the Sun, then it is even better as they will likely see repeat transits (shorter periods) and probably multiple transits (trend of systems being packed?).

    Probably you know this all and I have missed something…

  • Paul Gilster April 9, 2013, 8:27

    tesh writes:

    How can they claim to end up with any sort of reliable, confirmable, data for planets with longer periods?

    I’m going to the source to try to get this figured out. Hope to have more soon.

  • ljk April 9, 2013, 9:36

    coolstar said on April 8, 2013 at 15:19:

    “The extra-galactic bullies aren’t going to let people interested in stars and planets get very much time on it [JWST], you can be sure of that!”

    There is a tradition among professional astronomers that the more remote a galaxy is, the more important it is to the field, especially if it has an obscure catalog name. Planets, moons, planetoids, and the like are just so much local riff-raff. And alien life… puleeze.

  • Tom Womack April 9, 2013, 9:59

    I think the idea is that they’re looking at stars which are bright enough to do the follow-up with Earth-based telescopes; the catalogue of stars where TESS has seen one transit will be much richer than the normal input catalogue for a radial-velocity survey.

    It looks as if TESS is doing a two-month stare at each of a dozen pointings, so any orbital period less than a month gets you two transits and you can use Earth resources to check the expected next one.

  • Keith Cooper April 9, 2013, 10:11

    The statement about finding planets with periods up to one year does make sense to me; in a two year mission, where you need to observe three transits to confirm, a year is the maximum period that you can observe if you are lucky and happen to catch the first transit right at the beginning of the mission. I’m less sure about the two-month periods. How long does it take to survey the entire sky before it starts to repeat observations? Or rather, for how long will it have any particular part of the sky within its vision? That has to limit the length of some of the orbital periods that they can measure.

  • ljk April 9, 2013, 14:24

    Revisiting exoplanets and dark matter

    Two of the biggest topics in astronomy today are the search for extrasolar planets and the composition of dark matter. Jeff Foust updates some recent reports on those topics with news on new exoplanet missions and results from an experiment on the ISS.

    Monday, April 8, 2013


  • coolstar April 9, 2013, 16:54

    @ljk As a professional astronomer (of the stellar variety) I know this all too well! It’s a distinctly American prejudice though as at least Europeans don’t suffer from it. There would be very little stellar astronomy done by Americans in the last two decades had not planets been found around some of the stars!

  • Wojciech J April 10, 2013, 15:04

    Good news, telescopes like Cheops and TESS should give us more needed knowledge about exoplanets. We should also not forget about numerous ground-based projects that will start working around 2020s like EELT. Of course this hopefully will give finally way to projects like SIM and later Life Imager. Personally I am hopeful that by mid 20s scientists will have a schedule of deploying first swarm hypertelescope.

  • John Walker May 15, 2013, 22:28

    I would also very much be interested in understanding TESS’s survey method and how planets with orbital periods longer than 27 days are to be documented during its survey. One earlier iteration of the Mission from 2008 (here: http://exep.jpl.nasa.gov/ExoplanetForumFiles/12-2_Lathem_Ricker_transits.pdf ) suggested 3 x 2000 sq degree coverage once every 96 hour orbit. And 600 second measurements. Why has this changed? The current parameters call for 1 x 2000 sq degree coverage every 27 days ( 2 x 13 days plus midsegment downlink) While I agree that restricting the search exclusively to “Earthlike” planets is too narrow, it would have been nice to see if there was anything out there similar to the KOIs thought to be in an HZ. Most or all of which have orbital periods if I am correct of between 3 and 6 months. I would have preferred if TESS had been configured with the same stare and step format but with a 4 fold increase in the stare phase. At the cost of only surveying 1/4 of the sky during its 2 year mission.

  • John Walker May 17, 2013, 20:52

    I am a noob here. But, I suspect no one is reading this anyway. Whatever, of course for orbital period determination 3 transits are desired not 1 as I sugested above. Thus, my observation config just tripled in duration leaving only 1/12 of the sky observed to that degree in 2 years. Apparently, however, the goal of TESS is to primarily simply register transits with follow up observations of candidate systems to determine orbital and other characteristics done by ground based telescopes. This dearth of detail from the TESS data means massive telescope time for the follow up. Their own estimates were of 1400 full telescope nights with a 0.5 to 1.0 m class telescope and 300 nights with 1 – 2 m class instruments for classification of the host stars and for radial velocity measurements with roughly 1 km/s precision. 400 nights with 2-4 m class instruments for increased precision in m/s. And then an undetermined number of telescope nights for spectroscopic analysis of the smaller bodies using the largest class instruments (4-10 m). (ref. http://cips.berkeley.edu/events/rocky-planets-class09/false_positives_earths_brown_latham.pdf ) Where is that amount of telescope time going to come from? Wouldn’t it be more sensible to reduce the scope and increase the depth as I suggested above. I would be interested to hear the opinion of anyone else on this . Don’t know where else to ask.