Enceladus: The Leaky Moon With Tiger Stripes Written By Russell Adam Webb “Confirmation that the chemical energy for life exists within the ocean of a small moon of Saturn is an important milestone in our search for habitable worlds beyond Earth. Although we can’t detect life yet, we’ve found that there’s a food source there for it. It would be like a candy store for microbes,”said Linda Spilker, Cassini project scientist at NASA’s Jet Propulsion Laboratory (JPL) in Pasadena, California.
Parallel trenches on the south pole of Enceladus have been labeled ‘tiger stripes’ by enthusiastic scientists, but could life exist beneath the tiger stripes of this leaky moon? For those not particularly interested in astronomy (how could you not be?), Enceladus is a small icy moon that orbits the gas giant, Saturn. It’s not exactly somewhere we could ever settle, as it’s colder than liquid nitrogen, even in the summer and has very little gravity. But, one thing that excites us about Enceladus is the fact that beneath the icy surface lies a liquid water ocean. Size Comparison. The tiger stripes are hundreds of Kilometres long, a few kilometres wide and they’re jagged and cracked. Sometimes, the contents below, consisting of gas, water, ice and organic compounds will escape through small cracks and holes, and into space where we can examine it. This is what is getting the attention of astronomers lately, as it could hold the key to finding life outside of Earth. Enceladus is not longer than Great Britain Enceladus isn’t the only moon thought to have an ocean lurking beneath its cold surface, as Europa and others also have water. But, Enceladus is, so far, the only place we can confirm the existence of hydrothermal vents spewing out hydrogen gas. On Earth, microbes thrive in this environment and if given enough time, can evolve into a full ecosystem. Cassini arrived in the neighbourhood of Saturn in 2005 and has given us an unprecedented look at the area and its inhabitants. It stumbled upon the tiger strips, named so because it looks as though a set of claws have raked the surface, allowing us to peek inside. Future tourism location?
Cassini found that the tiger stripes were the warmest location of any of Saturn’s moons, or least cold we should say. A Lander could visit one day, but it wouldn’t be a very hospitable voyage if humans went along. If a Lander did visit this region, and burrowed down into the ice, it would find caverns filled with high pressure water vapor. If it burrowed further still, it would reach the haloed ocean. Cassini has been instrumental in deciphering what is going on beneath the surface of Enceladus, and astronomers think they have it all figured out. Dione, a neighbouring moon of, pushes the orbit of Enceladus into an ellipse. Heavy gravitational tides also act strongly on Enceladus, both ensuring the circular path and heating up the interior of the moon. Mysteries Remain There are still some mysteries we haven’t answered about everyone’s favourite new moon. First and foremost, we don’t know if life exists beneath the surface; it’s entirely possible if these circumstances have been around for enough time. Secondly, the fractures that the plumes escape from should freeze shut in a short period of time but these that we are studying now have been open for more than fifty years. We are also seeing fewer and fewer plumes and we are not sure why. They are now half as frequent as they were when Cassini arrived. We expect this to be a seasonal drop in volume, but we don’t know when they will start to increase again. Time is of the essence for Cassini, as the spacecraft that has captured the imagination of astronomy lovers across the world is almost out of fuel. Without the fuel, it won’t be able to make necessary course adjustments and stop itself from crash-landing somewhere. To avoid any possible contamination of places like Enceladus, NASA have decided to spectacularly crash the probe into Saturn’s surface.
Edited By Matt Dibble
The Breakthrough Starshot project achieved headlines recently, when they released detailed plans of a way to send a flotilla of tiny probes at extremely high speeds to reach Alpha Centauri in just 20 years. The creators of the project say that they can use lasers to speed up the sales on these small ships to travel at a fifth of the speed of light. The technology is existing, although would take years to develop into a working prototype. It would get us to our first exoplanet, Proxima B, and answer some important questions that we have about other stars. This estimate of 20 years involves a flyby. The probes would fly past Alpha Centauri at incredible speeds and take a very quick snapshot of the star. A flyby isn’t very valuable as in order to study the star, we need to slow down and take a closer look at it. René Heller and other independent researchers think that the light from a star can be used to decelerate the spacecraft via the solar powered sails. The math involved told the researchers that such a mission to Alpha Centauri would take around 140 years. The same math told the researchers that Sirius, a star that is twice the distance but 16 times as bright, could speed up and deceleratethe sails and get the spacecraft to the planet in just 69 years. Sirius is a nearby star that is twice as far as our closest neighbour, yet 16 times as bright. Researchers think that we can get there and study the star within 69 years. Astronomy has always felt fleetingly futile for me. Despite longing to know more, and yearning for the answer to the fundamental question of whether life exists elsewhere, everything in space is just so unimaginably far away. It might be my generation, or maybe it’s my own distinct lack of patience, but I want the really cool stuff to happen right now. I want probes to leave our solar system, I want drills to dig beneath the icy surface of moons like Enceladus and Europa and I want colonies living and thriving on the surface of Mars. The more I learn though, the more I become disillusioned as it all seems to take so impossibly long to make tangible progress. I’ve often thought that it would take an impossible and miraculous breakthrough with concepts such as faster-than-light travel to get anywhere. One German scientist, René Heller, at the Max Planck Institute for Solar System Research in Gottingen, Germany, thinks that we can get to Sirius in just 69 years. Sirius is not the closest star to Earth, as that title belongs to Alpha Centauri, but it is the brightest. Sirius is in fact, twice as far from our sun as Proxima Centauri, so how could we possibly get there quicker?
“We need a very light, solid, temperature-resistant, and highly reflective sail material that can span an area of several hundred meters squared. The material could possibly be based on graphene with a metamaterial coating, If this works out, then humanity can really go interstellar.” Ok so we might not be jetting to other star systems at warp speed anytime soon, and this might take a lifetime to achieve, but it makes me feel a little bit happier knowing that just maybe, we might actually travel to places that are incredibly far away whilst I’m still alive. Until then, we have the James Webb Space Telescope going up next year which should tell us a heap about the universe and how it was created. We can expect a million exoplanet discoveries, each one getting us more excited about the possibility of life outside our solar system and we can expect Edited By Matt Dibble
Astronomy Things To See During May 2017 (For UK Observers) Moon: First Quarter: Last Quarter: New:
3rd May, 3:47pm Full: 10th May, 10:43pm 19th May, 1:33am 25th May, 8:44pm
The Lunar “X” and “V” are visible at around 10am UT (11am BST) which is 90 minutes before the Moon rises so we can’t observe them from the UK this month
Lunar conjunctions & occultations: Note: When the Moon is waxing it is visible in the western sky after sunset. When near Full Moon it is visible most of the night. When it is waning, it is visible in the eastern sky before sunrise 1st May Waxing Crescent Moon lies between Pollux and Procyon 2nd May Nearly First Quarter Moon lies close to M44 the Beehive Cluster 3rd & 4th May First Quarter Moon lies close to Regulus 4th / 5th May Waxing Gibbous Moon occults 49 Leonis (see below) th 7 May Waxing Gibbous Moon lies extremely close to Jupiter 8th May Waxing Gibbous Moon lies close to Spica 10th May Full Moon lies close to Alpha & Beta Librae 11th May Waning Gibbous Moon lies close to Beta Scorpii th 12 May Waning Gibbous Moon lies close to Antares 14th May Waning Gibbous Moon lies close to Saturn & M23 cluster th 16 May Waning Gibbous Moon lies close to Pluto 17th May Waning Gibbous Moon forms a line with Alpha & Beta Capricorni th 20 May Thick Waning Crescent Moon forms triangle with Neptune & Lambda Aquarii 22nd May Waning Crescent Moon lies close to Venus 30th May Waxing Crescent Moon lies below M44 the Beehive Cluster 31st May Thick Waxing Crescent Moon lies close to Regulus
Planetary Observations: Mercury – is not observable this month Venus – is a dawn object this month and can be seen in the east before sunrise, rising about 20 minutes before the Sun by the end of the month. Shining at mag -4.4, it will be unmistakable! Mars – is visible low in the west after sunset for the first half of May, when it will be mag +1.6, then it will be lost in the evening twilight. During the first week of May it lies close to the Hyades cluster in Taurus. On 5th May, Mars lies just 6 degrees from Aldebaran, and on 27th May it lies close to the Waxing Crescent Moon Jupiter – located in Virgo, Jupiter shines brightly at mag -2.2 this month and is visible all night long. On 7th May, look for the extremely close conjunction of Jupiter and the Waxing Gibbous Moon, when they will be just 90 arcseconds apart. On 15th May, look for the 4 Galilean moons strung out in a line on the same side of Jupiter. With Jupiter so well placed, there are countless transit and shadow transit events involving the 4 Galilean moons during May. However, there is a double shadow transit of Io and Ganymede visible on the night of 27th /28th May, during which time Europa is occulted and reappears too. This all takes place between 19:13 BST and 03:26 BST Saturn – starting the month in Sagittarius and moving into Ophiuchus, mag +0.2 Saturn rises at around 11pm. On 9th May, see if you can spot its two-tone moon, Iapetus. It will have its bright hemisphere pointing towards us meaning that the moon will be 2 magnitudes brighter than usual. At mag +10.2 you need good binoculars or a telescope to spot it Neptune – located in Aquarius, Neptune is not visible at the beginning of May, but by the end of the month it will reappear in the morning sky, when it rises at around 2:30am. At mag +7.9 you will need binoculars or a small telescope to spot it. On 20th May Neptune lies close to the Waning Crescent Moon
Uranus – is not observable this month Pluto – located in Sagittarius, Pluto rises at around 1:30pm in the south east and remains visible until twilight. At mag +14.2 you will need a large telescope to spot it. On 16th May, Pluto lies close to the Waning Gibbous Moon Ceres – located on the edge of Taurus, Ceres moves ever closer to the Sun during the first half of May so it will be very difficult to observe this month Vesta – located in Cancer, Vesta becomes visible after sunset in the western sky, before setting at around 2am in the north west. At mag +7.4, you will need binoculars or a small telescope to spot it
Other Observations:. 49 Leonis Lunar Occultation –49 Leonis is occulted by the 8 day old Gibbous Moon overnight on 4th /5th May. 49 Leonis is a double star, containing a mag +5.8 star with a +7.9 companion which is 2 arcseconds from its companion. The pair will disappear behind the shadow side of the Moon at 00:20 BST and they re-emerge from the illuminated side around an hour later Eta Aquarid Meteor Shower – caused by passing through the debris stream of 2/P Halley, this shower peaks overnight on 5th/6th May. The radiant of the shower is close to the “Steering Wheel” asterism within Aquarius. This shower has a zenith hourly rate of around 50 meteors, but unfortunately the Waxing Gibbous Moon will severely hamper observations Noctilucent Cloud Season is Here! – May is the start of the northern hemisphere noctilucent cloud season. Although the peak isn’t until June/July, you may start to see NLCs during the last 2 weeks of May. At an altitude of around 8 times higher than other clouds, they are located on the edge of space. They are the edge of polar stratospheric clouds which are believed to be seeded by meteor dust. They can sometimes be seen around 60 – 120 minutes after sunset in the north west or 60 – 120 minutes before sunrise in the north east, but only between the end of May and mid August. They appear to glow a gorgeous white/blue whilst all the other clouds are in shadow, giving them their name “night shining clouds”. They are unpredictable, but if you get a good display, you will agree that they are well worth staying up late or getting up early for! Binocular Tour – This month’s Sky at Night Binocular Tour by Stephen Tonkin is focused on the sky around Coma Berenices and Virgo. There are 4 targets for 10x50 binoculars. First is the open cluster Merlotte 111. This misty patch will resolve into around 30 stars with binoculars, and contained within them is the double star 17 Comae Berenices. Next is a semi-regular variable star FS Comae Berenices. Its magnitude varies from +6.1 to +5.3 over a period of about 55 days. Next is the mag +7.7 globular cluster M53. Finally is the double star 32/33 Comae Berenices. There are 2 targets for 15x70 binoculars. First is the mag +8.5 galaxy M64, the Black Eye Galaxy. You will need good seeing conditions and no Moon to view this one. Finally is Markarian’s Chain, a chain of galaxies located in Virgo. You should be able to pick out at least 7 galaxies in this galaxy rich region. For full details on how to find these objects, look at this month’s edition of Sky at Night Magazine Deep Sky Tour – This month’s Sky at Night Deep Sky Tour is centred on the area around Cepheus. First is a target for a small telescope and that is Mu Cephei, which is a semi-regular variable red giant star, whose deep orange colour really stands out at low magnification. Its magnitude varies from +3.4 to +5.1. The next target for a small telescope is the mag +7.7 star cluster NGC 7235. This cluster contains about 100 stars, but a 10” telescope will only resolve around 15 of them. The next target is IC 1396, a diffuse nebula which although listed as mag +3.5, it actually has a very low surface brightness, making it appear fainter. A 6” telescope will show part of the edge of the nebula, but a 10” will show more of the eastern region. A UHC filter will help with this challenging target. There are 3 final targets for large telescopes, the first being IC 1396 the Elephant’s Trunk Nebula. Only a telescope of 14”+ and dark skies will reveal the bridge of the elephant’s nose. Finally there are 2 planetary nebulae to find. NGC 7008 the Foetus Nebula is a mag +12 planetary nebula with a mag +13.2 central star. High magnification will produce a lot of detail. The final target, Preite-Martinex 1-333 is a much more challenging planetary nebula. This was only confirmed as a true planetary nebula in 2009, and at mag +14 you will need at least a 16” telescope to see it. For full details of where to find these objects and how best to see them, pick up the current issue of Sky at Night magazine M5 (NGC 5904) – Astronomy Now’s object of the month is M5, a mag +5.7 globular cluster in Serpens. It is often overlooked in favour of M13, but is a very pretty cluster which is home to between 100,000 – 500,000 stars. From a dark sky site and with no Moon, it is just naked eye visible, however, it is an easy target in binoculars. To image this object you can use DSLRs or CCD cameras with LRGB filters. There is no benefit from using narrowband filters on globular clusters. For more information on how to observe, image or sketch this object, take a look at the current edition of Astronomy Now magazine
Solar Observations – the lengthening days this month give us more opportunity to observe the Sun. A white light filter will show sunspots, faculae and maybe some granulation. A specialist hydrogen-alpha telescope will show filaments, prominences and if you are lucky you may catch a solar flare in action. Also, if there is a lot of high level cirrus cloud around, keep a look out for solar optical phenomena such as parhelia (sundogs), 22 degree haloes and the various arcs associated with ice haloes SAFETY WARNING: Never attempt to observe or photograph the Sun without the correct equipment. Failure to do so will result in permanent damage to your eyes or even blindness! International Space Station – The ISS returns to UK skies during the 2nd week of May for some passes during the early hours of the morning. By the final week of May, there will be some late evening passes as well as the early morning ones. For the exact timings of the passes from your location, visit www.heavens-above.com You can also check the Iridium flare times for your location at Heavens Above
Comets Visible This Month: Comet C/2015 ER61 (PanSTARRS) – Located in Pisces, you may be able to spot this comet very low in the east before dawn as it reaches peak magnitude this month. Last reported visual observation of this comet was mag +6.5. During the 2nd week of May, it lies close to Venus. Click here to view the finder chart: http://bit.ly/2pWdnn4 Comet C/2015 V2 Johnson – moving from Hercules into Boӧtes, this comet is circumpolar and therefore visible all night long. The last reported observation put it at mag +7.7 but it is predicted to brighten during the month until it reaches peak magnitude at the end of May. Click here to view the finder chart: http://bit.ly/2oDDBc0 Comet C/2017 E4 Lovejoy – moving through Triangulum during May, this comet rises at around 3am in north east and remains visible until it is lost in the dawn twilight. The last reported visual observation put the comet at mag +7.5. Click here to view the finder chart: http://bit.ly/2oY2d2z Comet 41P/Tuttle-Giacobini-Kresak – moving through Hercules and heading towards Ophiuchus by the end of May, this comet is also circumpolar and visible all night long. The last reported visual observation put the comet at mag +7.5 Click here to view the finder chart: http://bit.ly/2lPvDhP There are several other comets in the mag +11 to +15 range. Details of these can be found in the links below. For up to date information about the fainter comets which are visible, please visit: https://in-the-sky.org/data/comets.php, the BAA Comets Section: https://www.ast.cam.ac.uk/~jds/ or Seiichi Yoshida’s home page: http://www.aerith.net/index.html
NB: All of the information in this sky guide is taken from Night Scenes 2017 by Paul L Money, Philips Stargazing 2017 by Heather Couper and Nigel Henbest, 2017 Yearbook of Astronomy by Richard Pearson and Brian Jones, Astronomy Now Magazine, Sky at Night Magazine, Stellarium, the BAA Comets Section website https://www.ast.cam.ac.uk/~jds/, www.inthesky.org and www.heavens-above.com Information collated by Mary McIntyre. For regular updates about the events happening in the sky this month, follow the Nightscenes Monthly Night Sky Facebook page at www.facebook.com/AstrospacePublications
A long held assumption is that stars which form within clusters do so from the same material at roughly the same time. That is to say, all the stars within a cluster are of similar age and similar composition. A star cluster is a group of stars that share a common origin and are held together by gravity for some length of time. Because star clusters are assumed to contain stars of similar age and composition, researchers have used them as an "astronomical laboratory" to understand how mass affects the evolution of stars. This assumption has been applied to current cluster models and is useful to test predictions regarding star mass variations. However, recent studies carried out by the International Centre for Radio Astronomy Research (ICRAR) located in Perth, Australia has cast doubts on this supposition. The ICRAR have studied several thousand stars located within 15 clusters inside the Large Magellanic Cloud and compared their cluster locations. Optical based telescopes are unable to penetrate the dust clouds within the densely packed clusters and so the ICRAR utilised both NASA’s Spitzar and ESA’s Herschel space telescopes to gather infra-red data. Research showed that 15 of the clusters coincided, whilst seven on the clusters contained young stars lying near to the centre of the cluster. The research showed that some nearby star clusters contain stars which are significantly younger than the rest of the cluster population. The formation of these younger stars could have been fuelled by gas entering the clusters from interstellar space. However, this theory was rejected by using observations made by radio telescopes to show that there was no correlation between interstellar hydrogen gas and the location of the clusters involved in the study. It is now proposed that these younger generational stars have possibly formed from ejected material from older stars within the cluster as they die. This proposal therefore supports the discovery of multiple generations of stars belonging to the same cluster.
Edited By Matt Dibble
Written By Andrew Richens, FRAS We know more about the Sun than any other star in the universe. Yet, new discoveries continue to unfurl regarding our closest stellar body. New research conducted by Durham University and NASA’s Goddard Space Flight Centre has developed a theoretical model to explain the trigger process which leads to solar eruptions. Until now, it was thought that different magnitudes of eruption were produced by different processes. However, the latest 3D computer modelling produced by the research suggests that the same trigger is responsible for solar eruptions ranging from the smallest coronal jet events (relatively small bursts of plasma) right up to Coronal Mass Ejections(CME’s). CME’s are massive clouds of plasma and magnetic fields which violently explode into space at tremendous velocities. Although it has previously known that both coronal jets and CME’s were linked to filaments – a twisting snake-like structure of dense plasma low within the suns atmosphere – it was unclear how their eruptions could result in such vastly different scales. Jets are triggered when the magnetic field lines above them break and then re-join. This process is known as magnetic reconnection. The new research suggests that Coronal Mass Ejections form in the same way, with the strength and structure of the magnetic field around the filament determining the type and magnitude of the eruption generated. The research has developed a new theoretical universal “break-out” model which predicts how stressed filaments push relentlessly at their magnetic restraints until they are able to break free of their shackles and escape into space. This model encompasses all coronal jets and CME’s, across all magnitudes of eruption. Understanding the trigger process of such violent solar events is of great importance. When Coronal Mass Ejection events are emitted in the direction of the Earth, charged particles accelerated by the CME collide with atoms within our atmosphere. This results in aurora; the Sothern and Northern Lights. However, not only can the movement of charged particles disrupt sensitive satellite equipment, the high-energy electrically charged particles can also endanger the lives of astronauts. Having a clearer understanding of the triggers for such events is therefore of great importance. Andrew Richens, FRAS
Edited By Matt Dibble
Written By Brian Jones
Today is the anniversary of the death, at Vienna on 2 May 1925, of the Austrian astronomer Johann Palisa. His main claim to fame is that he is the most successful visual discoverer of asteroids with a total of 122 to his name, all by the visual method rather than photography. The first of these was 136 Austria which he found on 18 Mar 1874, this during his term as Director of the Austrian Naval Observatory at Pola, Croatia (1872 to 1880) from where he made a total of 28 discoveries. In 1880 he moved to the newly-built Vienna Observatory from where he located a further 94, his last being of 1073 Gellivara on 14 Sep 1923. The 33km-diameter disintegrated lunar crater Palisa, located on the north western edge of Mare Nubium, is named in his honour. Edited By Matt Dibble
User Review by Paul Dibble As someone who is new to the world of astronomy and as such have never used any type of equipment for observing. The binoculars were very simple to un box and set up. The instructions provided by Meade were in plain English and easy to follow for someone who doesn’t yet fully understand Astronomy. After setting up the binoculars I was then able to view objects around me very easily, after night time fell I was able to view the night sky. Now I couldn’t tell you what it was that I could observe, however these binoculars gave me the opportunity to see stars, Planets and Galaxies of which I have never really seen before. The binoculars are powerful enough to give me a good field of view with its 4.4’ field of view. And given the Binoculars have a objective lens of 70mm and Magnification of x15 this makes for a fantastic viewing experience.
The Meade Astro-binoculars are a set of binoculars that are not only deigned for Astronomy but also for Nature and viewing object over a large distances. These Binoculars are designed to help amateurs understand more about astronomy with out the need to buy large Telescopes. The Binoculars are easy and simple to set up and use so much that even my young primary school aged children were able to use them and see into space. The Meade Astro-binoculars can be picked up for around £95 from the likes of Amazon, Tesco and Argos. Which is around the same costs as a Telescope, however for value for money the Astro-Binoculars are better as for the same observation quality and much more expensive telescope would be needed.
This telescope is perfect for children to learn and appreciate astronomy and to help adults view with ease. Given the Telescopes size it has given the children a much easier way to observe to universe and can be done in the comfort of our own home. However due to differentiating weather patterns and light pollution it can sometimes be hard to get out and use the telescope. The Celestron Telescope firstscope has been designed more for children to use. And as a basic telescope costs around ÂŁ50 to buy from various retail outlets. The Telescope specifications Size - 70mm Power - 15x & 70x Weight - 69oz.
So all in all a great Telescope and I would give the telescope Score 8/10
LETTER 1
doi:10.1038/nature22055
A temperate rocky super-Earth transiting a nearby cool star Jason A. Dittmann1, Jonathan M. Irwin1, David Charbonneau1, Xavier Bonfils2,3, Nicola Astudillo-Defru4, Raphaë lle D. Haywood1, Zachory K. Berta-Thompson5, Elisabeth R. Newton6, Joseph E. Rodriguez1, Jennifer G. Winters1, Thiam-Guan Tan7, Jose-Manuel Almenara2,3,4, Franç ois Bouchy8, Xavier Delfosse2,3, Thierry Forveille2,3, Christophe Lovis4, Felipe Murgas2,3,9, Francesco Pepe4, Nuno C. Santos10,11, Stephane Udry4, Anaë l Wü nsche2,3, Gilbert A. Esquerdo1, David W. Latham1 & Courtney D. Dressing12 Ks magnitude15 and empirically determined stellar relationships16,17, we estimate the stellar mass to be 14.6% thatof the Sun and the stellar M dwarf stars, which have masses less than 60 per cent radius to be 18.6% thatof the Sun. We estimate the metal contentof the that of the Sun, make up 75 per cent of the population of star to be approximately half thatof the Sun ([Fe/H] = −0.24 ± 0.10; all the stars in the Galaxy1. The atmospheres of orbiting errors given in the textare 1σ), and we measure the rotational period Earth-sized planets are observationally accessible via of the star to be 131 days from our long-term photometric monitoring transmission spectroscopy when the planets pass in (see Methods). On 15 September 2014 ut, MEarth-South identified a potential front of these stars2,3. Statistical results suggest that the nearest transiting Earth-sized planet in the liquid-water, transitin progress around LHS 1140, and automatically commenced habitable zone of an M dwarf star is probably around 10.5 high-cadence follow-up observations (see Extended Data Fig. 1). parsecs away4. A temperate planet has been discovered Using a machine-learning approach (see Methods), we selected this orbiting Proxima Centauri, the closest M dwarf5, but it star for further follow-up observations. We gathered two high-resolution (resolution R = 44,000) reconprobably does not transit and its true mass is unknown. Seven Earth-sized planets transit the very low-mass star naissance spectra with the TillinghastReflector Echelle Spectrograph TRAPPIST-1, which is 12 parsecs away6,7, but their masses (TRES) on the 1.5-m Tillinghast reflector located at the FLWO on and, particularly, their densities are poorly constrained. Here MtHopkins, Arizona, USA. From these spectra, we ruled outconwe report observations of LHS 1140b, a planet with a radius tamination from additional stars and large systemic accelerations, and of 1.4 Earth radii transiting a small, cool star (LHS 1140) 12 concluded thatthis system was probably nota stellar binary or false parsecs away. We measure the mass of the planet to be positive (see Methods). We subsequently obtained 144 precise 6.6 times that of Earth, consistent with a rocky bulk radial velocity measurements with the High Accuracy Radial Velocity composition. LHS 1140b receives an insolation of 0.46 Planet Searcher (HARPS) spectrograph18 from 23 November times that of Earth, placing it within the liquid- water, 2015 to 13 December 2016 ut. On 19 June 2016 ut, MEarth-South detected an additional transitof habitable zone8. With 90 per cent confidence, we place an upper limit on the orbital eccentricity of 0.29. The LHS 1140b through the MEarth trigger; when combined with the circular orbit is unlikely to be the result of tides and radial velocities and our inital trigger, we identified three potential orbital therefore was probably present at formation. Given its periods. On the basis of one of these, the 24.738-day period, we large surface gravity and cool insolation, the planet may back-predicted a third, low-signal-to-noise transit from 23 December have retained its atmosphere despite the greater 2014 ut (see Extended Data Fig. 1). With this ephemeris, we predicted luminosity (compared to the present-day) of its host star a transit on 01 September 2016 ut, the egress of which was in its youth9,10. Because LHS 1140 is nearby, telescopes observed by the Perth Exoplanet Survey Telescope (PEST; currently under construction might be able to search for see Methods). On 25 September 2016 and 20 October 2016 ut we obtained two complete transitobservations with four of the eight specific atmospheric gases in the future2,3. MEarth11,12 consists of two arrays of eight 40-cm-aperture tele- MEarth-South telescopes. We initally fited our photometric transit and radial-velocity measscopes, one in the Northern Hemisphere at the Fred Lawrence Whipple Observatory (FLWO) in Arizona, USA, and the other in the urements simultaneously (see Methods). This fit serves as input to Southern Hemisphere atCerro Tololo Inter-American Observatory, a comprehensive radial-velocity analysis thattakes into accountnotonly Chile. This survey monitors small stars (less than 33% the size of the the reflex motion of the planet, butalso the intrinsic variations of the host Sun) that are estimated to lie within 100 lightyears of the Sun for transiting star via Gaussian process regression19,20 (see Methods). We find extrasolar planets. Since January 2014, these telescopes have thatLHS 1140b has a mass 6.65 ± 1.82 times thatof Earth and aradius gathered data nearly every clear night, monitoring the brightnesses of 1.43 ± 0.10 times that of Earth, and orbits around LHS 1140 with a these stars for signs of slight dimming, which would be indicative of period of 24.73712 ±0.00025 days and an eccentricity that is a planettransiting in frontof the star. MEarth-South monitors these stars constrained to be less than 0.29 (at90% confidence; see Methods). 1 Harvard Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, Massachusetts 02138, USA. 2CNRS (Centre National de la Recherche Scientifique), IPAG (Institut de Planétologie et d’Astrophysique de Grenoble), F-38000 Grenoble, France. 3UniversitéGrenoble Alpes, IPAG, F-38000 Grenoble, France. 4Observatoire de Genève, Universitéde Genève, 51 chemin des Maillettes, 1290 Versoix, Switzerland. 5University of Colorado, 391 UCB, 2000 Colorado Avenue, Boulder, Colorado 80305, USA. 6Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, Massachusetts 02138, USA. 7Perth Exoplanet Survey Telescope, Perth, Western Australia, Australia. 8Aix Marseille Université, CNRS, LAM (Laboratoire d’Astrophysique de Marseille) UMR 7326, 13388 Marseille, France. 9Instituto de Astrofísica de Canarias (IAC), E-38205 La Laguna, Tenerife, Spain. 10Instituto de Astrofísica e Ciéncias do Espaço, Universidade do Porto, CAUP (Centro de Astrofísica da Universidade do Porto), Rua das Estrelas, 4150-762 Porto, Portugal. 11Departamento de Física e Astronomia, Faculdade de Ciências, Universidade do Porto, Rua do Campo, Alegre, 4169-007 Porto, Portugal. 12 Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, California 91125, USA.
00 MONTH 2017 | VOL 000 | NATURE
RESEA LETT a
b
1.004
2
R a d ia l v e lo c ity (m s –
1.002
R elative flux
1.000 0.998 0.996 0.994
1
0
–1
0.992 –2 0.990
–2 –1 0 1 Phased tim efrom mid-transit(h)
2
Figure 1 | Photometric transit and radial-velocity measurements of LHS 1140b. a, Phase-folded transitobservations from all transits (purple), with our transitmodel over-plotted as a red line. These data were binned in 160 3-min bins. Here we have corrected each individual light curve from each telescope with a zero-pointoffset, and with a linear correction for the air mass of the observation. For both full-transit observations we also apply a correction thatis linear in time so thatthe flux level of LHS 1140 is equal before and after the transit. b, 144
0 .0
0 .5 O rb ita l p h a s e
1 .0
(radial) velocity of LHS 1140 taken with the HARPS spectrograph (purple points, duplicate observations are shown in the shaded regions, error bars are 1σ). A value of zero corresponds to a radial velocity equal to thatof the hoststar. We have removed variability due to stellar activity and plotonly the radial-velocity perturbations induced by the planet, and have phase- folded the radial velocities to the orbital period. Our best-fiting Keplerian orbitis shown as the solid purple line.
Fig. 1, we show the phased and binned transitlightcurve for LHS 1140b Table 1 | System Parameters for LHS 1140 and the phased radial-velocity curve; in Extended Data Fig. 1 we show Parameter Value (median± 1σ) each individual transit. Our phased radial-velocity curve shows the Stellar parameters radial velocity from the influence of LHS 1140b only; the stellar con- RA (J2000) 00 h44 min59.3 s tribution has been removed. In Table 1, we show the system Dec. (J2000) −15°16′18″ parameters for LHS 1140 and LHS 1140b, with 68% confidence Proper motion 665.9± 1.0masyr −1 Position intervals on these parameters. 153.3°± 0.16°Photometry A simple structural model21 consisting of a dense iron core sur- angle V= 14.18 ± 0.03,R = 12.88 ± 0.02, rounded by a magnesium silicate mantle can explain the observed mass Ic= 11.19 ± 0.02,J = 9.612 ± 0.023, and radius (see Fig. 2). Although our bestfiting values imply a much H = 9.092 ± 0.026,Ks= 8.821 ± 0.024, W1= higher iron core-mass fraction than that of Earth (0.7 compared to 0.3), 8.606± 0.022, W2= 8.391± 0.019, our uncertainties on the mass and the radius can only rule outEarthW3= 8.235± 0.022, like compositions at 2σ confidence. We conclude thatLHS 1140b is W4= 8.133± 0.265 arocky planetwithouta substantial gas envelope. Distance to star, D* 12.47± 0.42parsecs We searched for additional planets both in the MEarth-South light (0.146± 0.019)M Mass, M* curve and as periodic signals in the HARPS residuals. We did not Radius,R* (0.186± 0.013)R find any compelling signals in these data. After subtracting the radial- Luminosity,L (0.002981 ± 0.00021)L * velocity signal due to stellar rotation and LHS 1140b, the Lomb–Scargle Effective temperature, T 3,131± 100K Metallicity, eff periodogram of the residuals shows a series of broad peaks at [Fe/H] −0.24± 0.10 (±0.1 periods greater than 60 days, possibly associated with stellar activity. At systematic) shorter periods, the highestpeaks are the resultof the window function Age, τ >5 Gyr * of our radial-velocity observations (see Methods). We do notfind any Rotational period, P 131 days rotation nota- ble periodic signals or additional triggers thatwould be suggestive Systemic velocity, γ −13.23± 0.60km s −1 * of another transiting planet, although planets smaller than LHS 1140b Transit and radial-velocity parameters could elude detection. 0.0708± 0.0013 Compact, coplanar, multi-planetsystems are common around M Planet-to-star radius ratio, Rp/R* 89.912°± 0.071° dwarfs22,23, and all coplanar objects atperiods of less than the orbital Inclination, i 5.34± 1.1m s −1 period of LHS 1140b would also transit, although their size may be Radial-velocity semi-amplitude, Kb <0.29 (90% confidence) too small to have been detected. The surface density of the protoplan- Eccentricity, eb Period, P 24.73712 ± 0.00025days Time etary disk in which LHS 1140b formed may dictate the properties of 2,456,915.6997 ± 0.0054HJD of mid-transit, T0 any additional planets in this system. A high-surface-density disk may 101.0± 6.0 Scaled orbital distance, a/R* shorten the timescale for planet formation to be before the dissipation of the gaseous disk component, allowing otherwise terrestrial plan- Derived planetary parameters (6.65± 1.82)M ets to accrete large envelopes of hydrogen and helium24. Mass, Mp (1.43± 0.10)R LHS 1140b does nothave such a gaseous component, so itmight Radius,Rp 12.5± 3.4g cm −3 formed from alow-surface-density disk, consistentwith the low metallicity Density, ρp 24.0± 2.7km s −1 of the star, although in this scenario we would expectadditional rocky Escape velocity, vesc Surface gravity, gp 31.8± 7.7m s −2 worlds. Further radial-velocity and photometric monitoring of Equilibrium temperature (0 albedo), Teq 230± 20K SemiLHS 1140 is warranted. major axis 0 . 0 8 7 5 ± 8 A recent study found that a planet orbiting an M dwarf could have 0.0041au surface temperatures thatallow liquid water if itreceives between 2 | NATURE | VOL 000 | 00 MONTH 20
RA, right ascension; dec., declination; M , R and L are the mass, radius and luminosity, respectively, of the Sun; M and R are the mass and radius, respectively, of Earth; HJD, heliocentric Julian date.
2
LETT RESEA 3.0
a
b
d
c
Planetary radius (R
2.5
2.0
LHS 1140b
1.5
LHS 1140b
LHS 1140b
LHS 1140b 1.0 Earth Venus 0.5
100101 Planetary mass (M )
104
103
102
101
10–
0.2 0.4 0.6 0.8 1.0 Stellar
101 102 103 Distance (parsecs)
radius (R )
Insolation (relative to Earth)
Figure 2 | Masses, radii, distances, insolation and stellar size of known transiting planets. a, The mass and radius of LHS 1140b indicate aterrestrial composition. Other planets with measured masses and radiiare shown, with darker points indicating smaller density uncertainties. The red points correspond to GJ 1214b (top) and GJ 1132b (below LHS 1140b). Error bars, 1σ. Mass–radius curves for two-layer rocky planets with 0%, 25% and 50% of their mass in iron cores are shown as solid lines. b–d, Planetary radius versus insolation, stellar radius and distance, respectively. Planets with dynamical mass determinations are shown in black; those withoutare shown in grey. The red data
to GJ 1214b and GJ 1132b, as in a, and the darker red circles to the TRAPPIST-1 planets; these are the nearby planets around small stars that are mostaccessible to characterization by the James Webb Space Telescope (JWST), as indicated in c and d. The shaded region in b is the M-dwarf habitable zone9. We note thatthis habitable zone is only appropriate for planets orbiting M dwarfs, and mostof the planets in this diagram orbitmuch larger stars. The area of each circle is proportional to the transitdepth and hence observational accessibility. LHS 1140b has a lower insolation than Earth, and orbits a small star 12 parsecs from the Sun, making ita temperate, rocky planetthatmay be accessible to atmospheric characterization.
0.2 and 0.8 times the insolation that Earth receives from the Sun. LHS 1140b currently receives 0.46 times Earth’s insolation, and we estimate its age to exceed 5 Gyr (see Methods). In its youth, LHS 1140 was more luminous, and a larger fraction of its spectrum was released at ultraviolet wavelengths. During this period, the atmosphere of LHS 1140b was therefore subjected to increased irradiance and greater levels of ionizing radiation, and LHS 1140b probably did notenter the liquid-water, habitable zone until approximately 40 Myr after the for- mation of the star9. This amountof time may have been sufficient for the atmosphere to have experienced a runaway greenhouse, with water being dissociated in the upper atmosphere and the hydrogen perma- nently lostto atmospheric escape9. If so, then the planet’s atmosphere would be dominated by abiotic O2, N2 and CO2. However, recent work has suggested that super-Earths can have an extended magma-ocean phase10, in which case the timescale over which LHS 1140b outgas- sed its secondary atmosphere may have exceeded the time for the star to reach its currentluminosity. In this scenario, volatiles such as H2O would have remained in the mantle of the planet until after the host star dimmed and its fractional ultraviolet emission decreased. Inferences of the history of the atmosphere would be strengthened with better observational constraints on the emission of M dwarfs atyoung ages, and with more detailed models of the inital composi- tion of the atmosphere from outgassing and the delivery of volatiles through late-stage cometary impacts. Observations of the currentultraviolet emission of LHS 1140 by the Hubble Space Telescope will be able to be used to assess the current high-energy flux that is infringing upon LHS 1140b, and will be helpful in determining the current habitability of LHS 1140b and constraining any ongoing atmospheric escape from the planet.
1.
Online Content Methods, along with any additional Extended Data display items and Source Data, are available in the online version of the paper; references unique to these sections appear only in the online paper.
2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
13. 14. 15. 16. 17. 18.
Henry, T. J. et al. The solar neighbourhood. XVII. Parallax results from the CTIOPI 0.9 m program: 20 new members of the RECONS 10 parsec sample. Astron. J. 132, 2360–2371 (2006). Rodler, F. & López-Morales, M. Feasibility studies for the detection of O2 in an Earthlike exoplanet. Astrophys. J. 781, 54 (2014). Snellen, I. A. G. et al. Finding extraterrestrial life using ground-based highdispersion spectroscopy. Astrophys. J. 764, 182 (2013). Dressing, C. D. & Charbonneau, D. The occurrence of potentially habitable planets orbiting M dwarfs estimated from the full Kepler dataset and an empirical measurement of the detection sensitivity. Astrophys. J. 807, 45 (2015). Anglada-Escudé, G. et al. A terrestrial planet candidate in a temperate orbit around Proxima Centauri. Nature 536, 437–440 (2016). Gillon, M. et al. Temperate Earth-sized planets transiting a nearby ultracool dwarf star. Nature 533, 221–224 (2016). Gillon, M. et al. Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature 542, 456–460 (2017). Kopparapu, R. K. et al. Habitable zones around main sequence stars: new estimates. Astrophys. J. 765, 131 (2013). Luger, R. & Barnes, R. Extreme water loss and abiotic O2 buildup on planets throughout the habitable zones of M dwarfs. Astrobiology 15, 119–143 (2015). Schaefer, L. et al. Predictions of the atmospheric composition of GJ 1132b. Astrophys. J. 829, 63 (2016). Nutzman, P. et al. Design considerations for a ground-based transit search for habitable planets orbiting M dwarfs. Publ. Astron. Soc. Pacif. 120, 317–327 (2008). Irwin, J. M. et al. The MEarth-North and MEarth-South transit surveys: searching for habitable super-Earth exoplanets around nearby M-dwarfs. In Proc. 18th Cambridge Workshop on Cool Stars, Stellar Systems, and the Sun (eds van Belle, G. & Harris, H. C.) 767–772 (2015). Berta, Z. K. et al. Transit detection in the MEarth survey of nearby M dwarfs: bridging the clean-first, search-later divide. Astron. J. 144, 145 (2012). Winters, J. G. et al. The solar neighbourhood. XXXVIII. Results from the CTIO/ SMARTS 0.9m: trigonometric parallaxes for 151 nearby M dwarf systems. Astron. J. 153, 14 (2017). Skrutskie, M. F. et al. The Two Micron All Sky Survey (2MASS). Astron. J. 131, 1163–1183 (2006). Benedict, G. F. et al. The solar neighbourhood. XXXVII. The mass-luminosity relation for main-sequence M dwarfs. Astron. J. 152, 141 (2016). Boyajian, T. S. et al. Stellar diameters and temperatures. II. Main-sequence Kand M-stars. Astrophys. J. 757, 112 (2012). Mayor, M. et al. Setting new standards with HARPS. The Messenger 114, 20–24 (2003).
00 MONTH 2017 | VOL 000 | NATURE
LETT RESEARCH 19. Rasmussen, C. E. & Williams, C. K. I. Gaussian Processes for Machine Learning Ch. 3 (MIT Press, 2006). 20. Haywood, R. D. et al. Planets and stellar activity: hide and seek in the CoRoT-7 system. Mon. Not. R. Astron. Soc. 443, 2517–2531 (2014). 21. Zeng, L. & Sasselov, D. D. A detailed model grid for solid planets from 0.1 through 100 Earth masses. Publ. Astron. Soc. Pacif. 125, 227 (2013). 22. Ballard, S. & Johnson, J. A. The Kepler dichotomy among the M dwarfs: half of systems contain five or more coplanar planets. Astrophys. J. 816, 66 (2016). 23. Muirhead, P. S. et al. Kepler-445, Kepler-446 and the occurrence of compact multiples orbiting mid-M dwarf stars. Astrophys. J. 801, 18 (2015). 24. Dawson, R. I. et al. Correlations between compositions and orbits established by the giant impact era. Astrophys. J. 822, 54 (2016). Supplementary Information is available in the online version of the paper. Acknowledgements We thank the staff at the Cerro Tololo Inter-American Observatory for assistance in the construction and operation of MEarth-South. The MEarth team acknowledges funding from the David and Lucille Packard Fellowship for Science and Engineering (awarded to D.C.). This material is based on work supported by the National Science Foundation under grants AST-0807690, AST-1109468, AST-1004488 (Alan T. Waterman Award) and AST-1616624. This publication was made possible through the support of a grant from the John Templeton Foundation and NASA XRP Program #NNX15AC90G. The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of the John Templeton Foundation. HARPS observations were made with European Southern Observatory (ESO) telescopes. This work was performed in part under contract with the Jet Propulsion Laboratory (JPL) funded by NASA through the Sagan Fellowship Program executed by the NASA Exoplanet Science Institute. E.R.N. is supported by an NSF Astronomy and Astrophysics Postdoctoral Fellowship under award AST-1602597. N.C.S. acknowledges support from Fundação para a Ciência e a Tecnologia (FCT) through national funds and by FEDER through COMPETE2020 by grants UID/FIS/04434/2013&POCI01-0145-FEDER-007672 and PTDC/FIS-AST/1526/2014&POCI-01-0145- FEDER016886. N.C.S. was also supported by FCT through Investigador FCT contract reference IF/00169/2012/CP0150/CT0002. X.B., X.D. and T.F. acknowledge the support of the INSU/PNP (Programme national de
4 | NATURE | VOL 000 | 00 MONTH 20
planétologie) and INSU/PNPS (Programme national de physique stellaire). X.B., J.M.A. and A.W. acknowledge funding from the European Research Council under ERC Grant Agreement no. 337591-ExTrA. We thank A. Vanderburg for backseat MCMCing. This publication makes use of data products from the Two Micron All Sky Survey (2MASS), which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by NASA and the National Science Foundation. This publication makes use of data products from the Wide-field Infrared Survey Explorer, which is a joint project of the University of California, Los Angeles, and the JPL/California Institute of Technology, funded by NASA. This research has made extensive use of the NASA Astrophysics Data System (ADS), and the SIMBAD database, operated at CDS, Strasbourg, France. Author Contributions The MEarth team (J.A.D., D.C., J.M.I., Z.K.B.-T., E.R.N., J.G.W. and J.E.R.) discovered the planet, organized the follow-up observations, and led the analysis and interpretation. J.A.D. analysed the light curve and the radialvelocity data and wrote the manuscript. J.M.I. designed and installed, and maintains and operates the MEarth-South telescope array, and contributed to the analysis and interpretation. D.C. leads the MEarth project, and assisted in analysis and writing the manuscript. E.R.N. determined the rotational period of the star. R.D.H. conducted the Gaussian process analysis of the radial velocities. J.E.R. and T.-G.T. organized the follow-up effort in Perth. The HARPS team (X.B., N.A.-D., J.-M.A., F.B., X.D., T.F., C.L., F.M., F.P., N.C.S., S.U. and A.W.) obtained spectra for Doppler velocimetry, with N.A.-D. and X.B. leading the analysis of those data. G.A.E. and D.W.L. obtained the reconnaissance spectrum with TRES at FLWO. C.D.D. obtained the infrared spectrum with IRTF/SpeX and determined the stellar metallicity. Author Information Reprints and permissions information is available at www.nature.com/reprints. The authors declare no competing financial interests. Readers are welcome to comment on the online version of the paper. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Correspondence and requests for materials should be addressed to J.A.D. (Jason.Dittmann@gmail.com). Reviewer Information Nature thanks A. Hatzes and the other anonymous reviewer(s) for their contribution to the peer review of this work.
LETT RESEA METHODS Mass of the star. Dynamical mass measurements of M-dwarf visual binaries, com- bined with accurate distance and magnitude measurements, have shown that there is a precise relationship between the near-infrared absolute magnitude of an M dwarf and its mass16,25. We used this relation16 to calculate the mass of LHS 1140 from its Ks magnitude, because the relationship in Ks is the mostprecise. We adopta mass of M*= (0.146 ±0.019)M , with the uncertainty of the mass primarily determined by the scatter in the mass–luminosity relation. Radius of the star. From this mass, we use a mass–radius relation calibrated with long-baseline optical interferometry of single stars17 to calculate the stellar radius: R* = (0.186 ±0.013)R . The Dartmouth Stellar Evolution Database26 offers a suite of stellar models across stellar mass, age and composition. Interpolating those models for a star of LHS 1140’s mass with [Fe/H] = 0 and [α/Fe] = 0 yields a stellar radius of (0.166 ± 0.11)R . Similarly, a model track with [Fe/H] = −0.5 and [α/Fe] =0 yields a stellar radius of (0.163 ± 0.010)R , consistent with the empirical relation thatwe have adopted. Additional stellar models27 atan age of 5.0 Gyr estimate a stellar radius of (0.166± 0.09)R , butunder-predictthe absolute Ks-band magnitude by 0.15 magnitudes. Inflating the model radius to correctthe absolute Ks magnitude to the observed value increases the stellar radius to 0.178R , consistent with the value thatwe adoptfor the system. The dominant source of error in these relations is our choice of the stellar mass; improvements in the determination of the stellar mass will improve the determination of the stellar radius accordingly. Bolometric luminosity of the star. We estimate the luminosity of LHS 1140 with the bolometric correction in V and in J from ref. 28. The absolute MJ magnitude and bolometric correction28 yield a luminosity of 0.00302L . Similarly, using the method of ref. 29, we find a luminosity of 0.002960L . We average these luminosities to estimate the stellar luminosity of LHS 1140 of L* = (0.002981 ± 0.00021)L . Using this luminosity estimate, we estimate the effective temperature of LHS 1140 to be Teff = 3,131± 100K. Identification of the initial trigger. The first observation of a transit of LHS 1140b by MEarth-South occurred in September 2014. However, we did notcommence follow-up until October 2015. The long period of LHS 1140b makes it difficult to identify via traditional techniques such as phase-folded searches, which is how MEarth identified GJ 1214b (ref. 30) and quickly confirmed GJ 1132b (ref. 31). Atthese long periods, single-longitude observatories can expect to be able to observe only one or two transits per season during night-time hours. This recovery rate is worsened by the likelihood of poor weather atthe site during any of these events. Therefore, itis unlikely that MEarth-South would ever have been able to identify LHS 1140b from phase-folded data alone. MEarth generates many triggers per night. The majority of these are false positives that can be explained through changes in the precipitable water vapour (which affects the differential brightness measurement between our ‘red’ target stars and our ‘blue’reference stars). MEarth corrects for this by measuring the change in relative flux of all of the M dwarfs stars thatare currently being observed, and constructing a ‘common-mode’vector of the light-curve flux behaviour, which is common to all targets. However, the time-resolution of this common mode is 30min, and more rapid changes in the precipitable water vapour atthe observing site are notcorrected and can potentially cause a MEarth trigger. To sortthrough and discard these triggers, we trained a neural network32 to classify MEarth data as either ‘trigger’data or ‘non-trigger’data, with the assump- tion that the majority of trigger data can serve as a proxy for data corrupted by systematic effects. We included MEarth triggers of GJ 1214 as non-trigger training events, because those observations were of a real transiting planetand notdue to systematic effects. Atthe time we executed this method, we had notyetconfirmed the existence of GJ 1132b, and therefore its triggers were omited from this analysis entirely (because we also could notjustify categorizing itas another false-positive). We included equal numbers of trigger observations and non-trigger MEarth observations in our training dataset. Non-trigger training data were selected such that, for each trigger thatan individual targetstar contributed to the trigger training set, a random non-trigger data point from that same star was selected to be included in the non-trigger training set. Future work will include the triggers of GJ 1132b and LHS 1140b and could improve the performance of this analysis. The nodes in our neural network are selected to be the weather and observatory state variables thatwe measure as partof routine MEarth observations. We selected the following state variables as the input nodes for our neural network: the atmos- pheric seeing, the ellipticity of the triggering image, the telescope pointing offsetfrom the field’s master image, the air mass, the zero-pointoffset(thatis, extinction from clouds), the root-mean-square variation of the reference stars, the common mode, its derivative, and its offset from a 3-h mean. The MEarth common mode is a measure of the average change in magnitude of all of the M dwarf stars thatMEarth is currently observing thatnight. This serves as a proxy
atmosphere. Including the derivative and the offsetof this common mode can help to determine whether the data point is potentially affected by rapidly changing atmospheric conditions. Our neural network is built to include one hidden layer of neurons, with the number of neurons in this hidden layer equal to the number of inputparameters (described above). The neural network is allowed to train and converge viathe back-propagation method. We did notinvestigate whether any of these input diagnostic parameters could be removed. We also did not investigate the effect of including differentnumbers of neurons in the hidden layer or including additional hidden layers. After training this network, we selected all of the trigger events (approximately 10% of all MEarth-South triggers) that were misclassified as non-trigger data for further scrutiny. We visually inspected each of these triggers, and selected the trigger eventfrom LHS 1140 to pursue. Photometric observations. With the exception of the 1 September 2016 transit, all observations were gathered with the MEarth-South telescope array. Both 2014 trig- ger events were observed by MEarth-South telescope 1. MEarth-South telescopes 1 and 6 observed the June 2016 trigger event. This event was observed through clouds. After the June 2016 trigger event, a combined analysis of the 15 September 2014 trigger, the 19 June 2016 trigger and the radial-velocity observations collected to thatdate revealed three candidate periods for LHS 1140b. One of these candidate periods also back-predicted a partial transitobserved by MEarth on 23 December 2014. We selected this ephemeris to predict future transit opportunities and we pursued four of these. We attempted to observe a full transitof LHS 1140b from Australia on 7 August 2016, but all sites were weathered out. We attempted to observe an additional transit from Australia and Hawaii on 1 September 2016. Five of the six sites were weathered outand we obtained a partial transit, includ- ing egress, from the Perth ExoplanetSurvey Telescope (PEST). We succeeded in obtaining two full transitobservations with MEarth-South on 25 September 2016 and 20 October 2016. The 25 September 2016 and the 20 October 2016 full-transit events were observed by MEarth-South telescopes 1, 2, 6 and 8. Exposure times were 23 s, yielding a cadence of 46 s for each of the MEarth-South telescopes for the high- cadence follow-up observations. PEST is a home observatory with a 12-inch Meade LX200 SCT f/10 telescope with a SBIG ST-8XME CCD camera. The observatory is owned and operated by Thiam-Guan (TG) Tan. PEST is equipped with a BVRI filter wheel, a focal reducer yielding f/5, and an Optec TCF-Si focuser controlled by the observatory computer. PEST has a 31′× 21′field-of-view and a 1.2″pixel scale. The observatory clock is synced on start-up to the atomic clock in Boulder, Colorado, and is re-synced every 3 h during an observing night. PEST observed an egress of LHS 1140b on 1 September 2016 in the I-band using 120-s exposures, and obtained a cadence of approximately 132s. We show each individual transitin Extended Data Fig. 1. Radial velocity observations. We first gathered reconnaissance spectra of LHS 1140 with the TRES spectrograph atFLWO. We measured a velocity shiftbetween two spectra of 37 ± 34 ms −1, which is consistent with 0 m s−1, and began observations with the HARPS spectrograph on the La Silla 3.6-m telescope. We gathered 144 observations between 23 November 2015 and 13 December 2016, using an exposure time of 30 min and collecting two back-to-back exposures, although on three nights only one exposure was obtained. Each spectrum spans from 380 nm to 680 nm in wavelength. For each observation, we constructed acomparison template by co-adding all of the other spectra and measured the rela- tive radial velocity as a shift required to minimize the χ2 of the difference between the spectrum and this template33,34. Telluric lines were masked using a template of lines made with a different, much larger dataset. We see no evidence for rotational broadening of the spectrum, consistentwith the measured rotational period of LHS 1140. The median of the estimated internal error of the radial velocities is 4.1m s−1 per observation. Simple simultaneous photometric and radial velocity analysis. We fit our lightcurves simultaneously with our radial velocities, letting the following parameters vary: orbital period P, time of mid-transitT0, planet-to-star radius ratio Rp/R*, radi- al-velocity semi-amplitude Kb, orbital inclination i, orbital eccentricity eb, argumentof periastron ωb, and separate baseline flux levels for each transitobserved on each telescope. We simultaneously correcteach lightcurve for a linear term in air mass to correct for differential colour extinction, and we also correct both full-transit observations with a linear trend in time, so thatthe out-of-transitbaseline is equal on both sides of the transit. Limb darkening was treated with a quadratic approximation, with coefficients
LETT RESEARCH limited amount of in-transit data for our PEST observation, we use the same limbdarkening parameters for all of our data. Our final limb-darkening parameters are a = 0.219± 0.025 and b = 0.415± 0.083. The light-curve model is generated from JKTEBOP36, modified to compute integrals analytically using the methods of ref. 37. We explore the parameter space with the emcee code38, which is a Python implementation of the affineinvariantMarkov chain Monte Carlo (MCMC) sampler, and take the 16th, 50th and 84th percentiles of the resultantparameters to obtain the 68% confidence interval for each of our model parameters. For the radial velocities we adopta range of functional forms to describe the radial-velocity signal caused by stellar activity, which we describe below. We show the Lomb–Scargle39,40 periodogram of the window function (the periodogram in which all radial-velocity measurements are setto 1) and of the measured HARPS radial velocities in Extended Data Fig. 2. Our window function contains substantial power atP = 18 days and atits harmonics, as well as near P = 40 days. Our measured radial velocities contain substantial power at the true orbital period (P= 24.73712 days) of LHS 1140b as well as power athigher peri- ods (90 and 130 days). This excess power at high periods is due to radial-velocity variations induced by the stellar activity of LHS 1140. Stellar activity is known to produce systematic radial-velocity signal due to convective inhibiton of starspots and to the flux imbalance between dark spots and brightfaculae41–43. We assume a simple functional form for the radial-velocity variations induced by stellar rotation— a single sine wave. We initalize this sinusoid atthe meas- ured photometric rotational period of LHS 1140 (131 days) and an amplitude of 3 ms −1, and allow these model parameters to vary freely. Our final best-fiting model has a radial-velocity rotational activity period of 129.5± 4.5 days, an amplitude of 4.26 ± 0.60m s−1 and an epoch of 2,457,641.1± 3.6 HJD. The period of this best-fiting sinusoid is consistent with the photometric rotational period. In Extended Data Fig. 3, we show the radial velocities of LHS 1140 phased to this period, along with our bestfit. When subtracting this activity-induced radial-velocity variation from our HARPS measurement, the Lomb–Scargle periodogram of our residual radial veloc- ites contains reduced noise at long periods. Additionally, the highest remaining peak in this periodogram is the orbital period of LHS 1140b (see Extended Data Fig. 4a). After subtracting our best-fiting Keplerian orbit for LHS 1140b from these residual data, the Lomb–Scargle periodogram shows no more large peaks (Extended Data Fig. 4b). The residual peak near P= 18 days is due to the window function of our observations. We see no evidence for additional radial-velocity variations due to additional planets orbiting LHS 1140b, although additional obser- vations and future work in which stellar velocity jiter due to activity is subtracted may improve the mass determination of LHS 1140b and potentially uncover additional signals from planets. Radial velocity observations can also show substantial power atthe harmonics of the stellar rotational period and impede planetdetection20. To investigate the effect thatthis could have on our fitwe performed an identical analysis, butalso included sinusoidal terms with periods of Protation/2 and Protation/3, where Protation is the stellar rotational period. The stellar rotational period is again initalized atthe photometric rotational period and allowed to vary slightly. We find that, in this scenario, the best-fiting semi-amplitude for the radial-velocity signal of the planethas an amplitude of 6.0± 0.7m s −1, slightly higher than, but consistent with, the singlesinusoid model. We also performed an identical fit, replacing the 131-day stellar rotational period and its harmonics with the 90-day peak seen in the radial-velocity periodogram and its harmonics. In this scenario, we find a semi-amplitude of 5.7± 0.7m s−1 for the radial-velocity variation induced by LHS 1140b, consistentwith our other fits. We note thatfor all of these analyses, we enforce a simple and strictfunctional form to describe the stellar contribution to the radial-velocity signal. In reality, this does not accurately model these effects, and so to obtain an uncertainty that more accurately reflects our uncertainties in the radial-velocity activity of the star, we turn to a more flexible model than we have presented here. Radial-velocity analysis with Gaussian process regression. Features on the sur- face of the host star, such as dark spots, faculae and granulation, produce signals modulated by the stellar rotational period and its firstand second harmonics20. These radial-velocity signals also evolve over time, and will thus produce qua- si-periodic signals with varying amplitudes and phases. Therefore, the sinusoidal models above do not represent a perfect description of the stellar-induced radi- al-velocity variation. Theoretical44,45 and observational20,46 studies suggest that the surfaces of M dwarfs are covered with large numbers of small, low-contrast spots, which produce radial-velocity variations similar in amplitude to planetary-induced orbital motion, probably through inhibiton of convection and flux imbalances. The exactshape of activity-induced radial-velocity
understood. We see no evidence of correlations between radial velocities and the spectroscopic activity indicators (Ca i H&K-derived S index and Hα emission) or between radial velocities and indicators of the asymmetry of the cross-correlation function (full-width at half-maximum and bisector span), although this is not unexpected in such a slowly rotating star41. Furthermore, the activity signals in radial velocities will notbe perfectly correlated with photometric variations20,47, which are sensitive to only the longitudinally asymmetric portion of the starspotdistribution. To accountfor activity-induced rotational modulation and evolution in the radial velocities we use Gaussian process regression19, a technique that has been applied successfully in previous radial-velocity mass determinations of planets around active stars such as CoRoT-720,48, Kepler-7849, Alpha CentauriB50 and Kepler-2151. Gaussian process regression allows us to incorporate the uncertainty that arises from magnetic activity of the star in our determination of the planetary mass, while making minimal assumptions about the exact form of activity-induced radial-velocity signals. Our Gaussian process has a quasi-periodic covariance kernel with timescales of recurrence R and evolution E that correspond to the rotational period and active-region lifetime of the star, respectively. We measure the rotational period from the light curve to be 131 ± 10 days (which is also consistent with the
′ -deri Rlog() Ris′ the chromospheric emission HK ved value of 127± 11days, whereHK ratio, defin ed as th e ratio of th e fluxes from th e Ca H and K lin es to the broadband photometric flux in the R band). The periodogram of the radial velocities (Extended Data Fig. 2b) displays a peak not only around 130 days, butalso at P= 90 days; however, this peak is present in the window function (Extended Data Fig. 2a), which means that it is probably a fake signal that arises from our observing strategy rather than a true stellar signal. To prevent hyperparameter R from settling at this 90-day value, we impose a Gaussian prior with a slightly narrower width of 5 days instead of 10 days. The lightcurve is too shortto give a precise estimate of the average lifetime of the active regions on the stellar surface, but we can infer that it is longer than a few rotational cycles; we adopt a value of three times the stellar rotational period (393 days). We constrain hyperparameter E with a Gaussian prior centred at this value, with a width of 30 days. We find that choosing a different value (of the same order of magnitude) does not greatly affectour mass determination and its associated uncertainty. Another parameter, S, determines the maximum number of peaks thatthe activity-induced radial-velocity signal is allowed to have within each recurrence timescale. Radial-velocity curves typically show two or three peaks per rotation (limb effects and degeneracies wash out small-structure signatures), which can be reproduced with S = 0.5. We are therefore justified in adopting a strong Gaussian prior for this parameter (0.5± 0.05). The amplitude A of the Gaussian process is treated as a free parameter, drawn from a uniform and positive prior distribution. The orbitof the planetis modelled as a Keplerian orbitwith free eccentricity, with orbital period and time of transitconstrained by Gaussian priors based on our transit analysis. The radial-velocity semiamplitude Kb, is drawn from a uniform, positive prior distribution. We account for the presence of additional uncorrelated, Gaussian noise with a term added in quadrature to the diagonal elements of the covariance matrix. These diagonal elements correspond to the variance of each observation, which are rou- tinely taken as the formal radial-velocity uncertainties. It is drawn from a Jeffreys prior distribution with a lower bound equal to the internal instrumental precision of HARPS (0.6m s−1). The best-fiting parameters and their associated uncertainties are determined through an MCMC procedure akin to that described in ref. 51. Our best-fiting radial-velocity model is shown in Extended Data Fig. 5 and Extended Data Fig. 6 shows correlation plots for all of the parameters of our radial-velocity model; no substantial correlations are present and all of the parameter chains achieve good convergence. We measure a radial-velocity semi-amplitude of Kb = 5.34 ± 1.1 m s−1 for LHS 1140b, which corresponds to a mass of Mp = (6.65± 1.82)M . We place an upper limit on the eccentricity of eb < 0.29 with 90% confidence. We adopt these values as our final values for the physical parameters of the LHS 1140 system. Rejection of false-positive scenarios. Owing to the high proper motion of LHS 1140, we are able to inspect the POSS-I E photographic plates for luminous background objects located atthe currentposition of LHS 1140. These plates have a magnitude limit of 19.5 in a band pass roughly similar to the R band, approximately 9 magnitudes fainter than LHS 1140. We see no source at the currentposition of LHS 1140, and therefore conclude thatthere is no background source contaminat- ing our transit measurements or producing a false-positive signal. To investigate the possibility of a luminous or massive bound companion in the LHS 1140 system, we inspected the high-resolution spectra taken with HARPS
LETT RESEA no evidence for a radial-velocity drift over these observations or any additional spectral features. On the basis of the lack of a reported perturbation on the parallax14 and the ten-year span of data, we can rule outany signal due to a brown dwarf companion. A bound L0 brown dwarf at 5 au would be approximately 6 magnitudes fainter in V, but would induce a perturbation of 175 mas, twice the parallax amplitude of LHS 1140. Given these strict astrometric limits on bound companions, we reject false-positive scenarios and substantial contamination by a third light source in our transitmeasurements. The eccentricity of LHS 1140b. We have ruled out eccentricites larger than eb = 0.29 for LHS 1140b, and our observations are consistentwith a circular orbit. Unlike the other small planets that orbit mid-to-late M dwarfs and have well-constrained eccentricites (GJ 121430 and GJ 113231), tidal circularization is inefficientatthe orbital distance of LHS 1140b, and its circular orbitis probably natal. This means that LHS 1140b either formed in its present location or, if it migrated, this migration musthave proceeded smoothly in the protoplanetary disk. More violentmigration mechanisms, such as planet–planetscattering, would have produced asystem with large orbital eccentricity. Even though we constrain the eccentricity of LHS 1140b to be low, future determination of its precise value is important for discussions of its climate and potential habitability. Metallicity of the star. We gathered a near-infrared spectrum of LHS 1140 with the IRTF/SPeX instrument52. We used an observing time of 60s, with the 0.3″× 15″ slitand a resolving power of R= 2,000. We used HD 3604 as our AOV standard star. We compared this spectrum by eye to the IRTF Spectral Library53,54 and classified LHS 1140 as an M4.5V-type star (Extended Data Fig. 7). We use near-infrared spectral features thathave been shown to be sensitive indicators of stellar metallicity55,56 to estimate the metallicity of LHS 1140. For features located in the H band, we find a value of [Fe/H] = −0.04 ± 0.10 and [M/H] = −0.24 ± 0.09; for features in the Ks band, we find [Fe/H] =−0.30 ± 0.08 and [M/H] = −0.25 ± 0.08. An estimate of the MEarth-band magnitude of LHS 1140, combined with our photometric metallicity relation57, yields an esti- mate of [Fe/H]= −0.22±0.1. All of these relations suffer from a relative lack of low-metallicity calibrator systems, and the uncertainty is probably higher than the internal errors reported for each index. However, with the exception of [Fe/H] indicators in the H band, we consistently find that LHS 1140 is a metal-poor star compared to the Sun, and we adopta metallicity of [Fe/H]= −0.24±0.10, where this error is the standard deviation of the individual estimates and does not accountfor any systematic errors internal to the calibrations themselves. We estimate this systematic error to be of the same magnitude as the scatter in the relation itself. Rotational period and age of the star. MEarth data are calibrated each nightusing observations of standard star fields. Analysis of these data58,59, calibrating each nightand correcting for the effects of camera changes and the slow loss of sensitivity from dust settling on the instruments, has allowed us to measure the photometric rotational periods for many of our targets. We measure a photomet- ric modulation, which we interpret as the rotational period, of 131 days for LHS 1140b (Extended Data Fig. 8). This rotational period is consistentthrough- outthe MEarth-South observat seasons ste± nt ′ ional (ref. 60) andofis consi 127 Rlog() HK bet een telescopes. We estimphot ate oamet rotraictiomeasurement nal perio d from 11wdays, consistent wi th our . We also find that the amplitude of the photometric modulation due to rotation has increased between the 2014– 2015 observing season and the 2016 observing season. Because this signal is determined by the longitudinal asymmetric distribution of starspots, it is difficultto tell whether this is an indication of a long-term trend, partof a stellar cycle, or simply due to a slightincrease in starspotasymmetry over the MEarth-South observational baseline. M dwarfs spin more slowly as they age, with lessmassive stars reaching longer rotational periods atolder ages than more massive stars61. Barnard’s star has a rotational period of 130 days (ref. 62), a mass of 0.16M and is estimated to be 7–13 Gyr old63. This suggests that LHS 1140 is likely to be comparable in age to Barnard’s star. LHS 1140 has a heliocentric (U, V, W) velocity63 of (3.64 ± 0.42, −40.77± 1.3 5, 5.40 ± 0.27) kms −1. The Sun’s motion relative to the local standard of rest .06 9 has been measured to111 be ( , 12.24 ± 0.47, 7.25 ± 0.37) km s −1 (ref. . − .+075 64). Correcting to the local standard of rest, we find thatLHS 1140 has a (U, V, W) . 08 0 velocity of ( − . − .74 , −53.01 ± 1.43, −1.85 ± 0.46) km s−1 and is consistent 6+086 with a kinematically older stellar population65, although this is not a tightconstraint. Furthermore, we detectno Hα emission in our spectra. M4.5V dwarfs tend to have active Hα emission for the first 5 Gyr of their lives66,67. On the basis of gyrochronology and asteroseismology relations68, the age of Alpha CentauriB is estimated to be approximately 6–7 Gyr. The rotational period of Proxima Centauri is approximately 86 days (ref. 69), suggesting that LHS 1140 may be older than
Proxima Centauri. Taken together, we suggestthatLHS 1140 is likely to be greater than 5 Gyr in age. Code availability. The EMCEE code is available as a Python install and is publicly available on GitHub (https://github.com/dfm/emcee). JKTEBOP and the modifi- cations performed to it are also publicly available on GitHub (https://github.com/ mdwarfgeek/eb). These codes were used to produce the lightcurve and radial- velocity models used in the analysis of our data. The code used to determine the planetary orbitand mass, with Gaussian process regression to account for the magnetic activity variations of the host star, is not yet publically available, butwe are currently working to make itso. Data availability. All data used in this work are provided as Supplementary Data, and are available on the MEarth project webpage (https://www.cfa.harvard.edu/ MEarth/Welcome.html) and via the online repository FigShare (https://figshare. com/s/9e2b29d4f7a8043ca071 for MEarth and PEST photometry; https://figshare. com/s/49625e95aabf9e1f2ae6 for HARPS radial-velocity data). 25. Delfosse, X. et al. Accurate masses of very low mass star. IV. Improved mass-luminosity relations. Astron. Astrophys. 364, 217–224 (2000). 26. Dotter, A. et al. The Dartmouth stellar evolution database. Astrophys. J. Suppl. Ser. 178, 89–101 (2008). 27. Baraffe, I., Homeier, D. Allard, F. & Chabrier, G. New evolutionary models for premain sequence and main sequence low-mass stars down to the hydrogenburning limit. Astron. Astrophys. 577, A42 (2015). 28. Pecaut, M. J. & Mamajek, E. E. Intrinsic colors, temperatures, and bolometric corrections of pre-main-sequence stars. Astrophys. J. Suppl. Ser. 208, 9 (2013). 29. Mann, A. W. et al. Spectro-thermometry of M dwarfs and their candidate planets: too hot, too cool, or just right? Astrophys. J. 779, 188 (2013). 30. Charbonneau, D. et al. A super-Earth transiting a nearby low-mass star. Nature 462, 891–894 (2009). 31. Berta-Thompson, Z. K. A rocky transiting planet transiting a nearby low-mass star. Nature 527, 204–207 (2015). 32. Haykin, S. Neural Networks: A Comprehensive Foundation (Prentice Hall, 1998). 33. Bouchy, F. et al. Fundamental photon noise limit to radial velocity measurements. Astron. Astrophys. 374, 733–739 (2001). 34. Astudillo-Defru, N. et al. The HARPS search for southern extra-solar planets. Astron. Astrophys. 575, A119 (2015). 35. Claret, A. et al. New limb-darkening coefficients for PHOENIX/1D model atmospheres. I. Calculations for 1500 K ≤ Teff ≤ 4800 K. Kepler, CoRoT, Spitzer, uvby, UVBRIJHK, Sloan, and 2MASS photometric systems. Astron. Astrophys. 546, A14 (2012). 36. Southworth, J. The solar-type eclipsing binary system LL Aquarii. Astron. Astrophys. 557, A119 (2013). 37. Mandel, K. & Agol, E. Analytic light curves for planetary transit searches. Astrophys. J. 580, L171–L175 (2002). 38. Foreman-Mackey, D. et al. emcee: the MCMC Hammer. Publ. Astron. Soc. Pacif. 125, 306 (2013). 39. Lomb, N. R. Least-squares frequency analysis of unequally spaced data. Astrophys. Space Sci. 39, 447–462 (1976). 40. Scargle, J. D. Studies in astronomical time series analysis. II –statistical aspects of spectral analysis of unevenly spaced data. Astrophys. J. 263, 835–853 (1982). 41. Desort, M. et al. Search for exoplanets with the radial-velocity technique: quantitative diagnostics of stellar activity. Astron. Astrophys. 473, 983–993 (2007). 42. Boisse, I. et al. Disentangling between stellar activity and planetary signals. Astron. Astrophys. 528, A4 (2011). 43. Haywood, R. D. et al. The Sun as a planet-host star: proxies from SDO images for HARPS radial-velocity variations. Mon. Not. R. Astron. Soc. 457, 3637–3651 (2016). 44. Meunier, N. et al. Using the Sun to estimate Earth-like planets detection capabilities. II Impact of plages. Astron. Astrophys. 512, A39 (2010). 45. Dumusque, X. et al. SOAP 2.0: a tool to estimate the photometric and radial-velocity variations induced by stellar spots and plages. Astrophys. J. 796, 132 (2014). 46. Meunier, N. et al. Reconstructing the solar integrated radial velocity using MDI/ SOHO. Astron. Astrophys. 519, A66 (2010). 47. Aigrain, S. et al. A simple method to estimate radial-velocity variations due to stellar activity using photometry. Mon. Not. R. Astron. Soc. 419, 3147–3158 (2012). 48. Faria, J. P. et al. Uncovering the planets and stellar activity of CoRoT-7 using only radial velocities. Astron. Astrophys. 588, A31 (2016). 49. Grunblatt, S. K. et al. Determining the mass of Kepler-78b with nonparametric Gaussian process estimation. Astrophys. J. 808, 127 (2015). 50. Rajpaul, V. et al. A Gaussian process framework for modelling stellar activity signals in radial-velocity data. Mon. Not. R. Astron. Soc. 452, 2269–2291 (2015). 51. López-Morales, M. et al. Kepler-21b: a rocky planet around a V = 8.25 magnitude star. Astron. J. 152, 204 (2016). 52. Rayner, J. T. et al. SpeX: a medium-resolution 0.8-5.5 micron spectrograph and imager for the NASA Infrared Telescope Facility. Publ. Astron. Soc. Pacif. 115, 362–382 (2003).
LETT RESEARCH 53. Rayner, J. T. et al. The Infrared Telescope Facility (IRTF) spectral library: cool stars. Astrophys. J. Suppl. Ser. 185, 289–432 (2009). 54. Cushing, M. C. et al. An infrared spectroscopic sequence of M, L, and T dwarfs. Astrophys. J. 623, 1115–1140 (2005). 55. Mann, A. W. et al. Prospecting in late-type dwarfs: a calibration of infrared and visible spectroscopic metallicities of late K and M dwarfs spanning 1.5 dex. Astron. J. 145, 52 (2013). 56. Mann, A. et al. How to constrain your M Dwarf: measuring effective temperature, bolometric luminosity, mass, and radius. Astrophys. J. 804, 64 (2015); erratum 819, 87 (2016). 57. Dittmann, J. A. et al. Calibration of the MEarth photometric system: optical magnitudes and photometric metallicity estimates for 1802 nearby M-dwarfs. Astrophys. J. 818, 153 (2016). 58. Irwin, J. et al. On the angular momentum evolution of fully-convective stars: rotation periods for field M-dwarfs from the MEarth transit survey. Astrophys. J. 727, 56 (2011). 59. Newton, E. R. et al. The rotation and galactic kinematics of mid M dwarfs in the solar neighborhood. Astrophys. J. 821, 93 (2016). 60. Astudillo-Defru, N. et al. Magnetic activity in the HARPS M-dwarf sample. The ′ . Astron. rotation-activity relationship for very low-mass stars through RHK Astrophys. 600, A13 (2017).
61. Benedict, G. F. et al. Photometry of Proxima Centauri and Barnard’s star using Hubble Space Telescope Fine Guidance Sensor 3: a search for periodic variations. Astron. J. 116, 429–439 (1998). 62. Feltzing, S. & Bensby, T. The Galactic stellar disc. Phys. Scr. T133, 014031 (2008). 63. Johnson, D. R. H. & Soderblom, D. R. Calculating galactic space velocities and their uncertainties, with an application to the Ursa Major group. Astron. J. 93, 864–867 (1987). 64. Schönrich, R. et al. Local kinematics and the local standard of rest. Mon. Not. R. Astron. Soc. 403, 1829–1833 (2010). 65. Bochanski, J. J. et al. Exploring the local Milky Way: M dwarfs as tracers of Galactic populations. Astron. J. 134, 2418–2429 (2007). 66. West, A. A. et al. Constraining the age-activity relation for cool stars: the Sloan Digital Sky Survey data release 5 low-mass star spectroscopic sample. Astron. J. 135, 785–795 (2008). 67. Newton, E. R. et al. The Hα emission of nearby M dwarfs and its relation to stellar rotation. Astrophys. J. 834, 85 (2017). 68. Mamajek, E. E. & Hillenbrand, L. A. Improved age estimation for solar-type dwarfs using activity-rotation diagnostics. Astrophys. J. 687, 1264–1293 (2008). 69. Collins, J. M., Jones, H. R. A. & Barnes, J. R. Calculations of periodicity from Hα profiles of Proxima Centarui. Astron. Astrophys. https://doi.org/10.1051/00046361/201628827 (2017).
LETT RESEA
Extended Data Figure 1 | Individual transit light curves observed with MEarth and PEST. The top three lightcurves are MEarth-South trigger observations whereas the bottom three are targeted transitobservations. Lightcurves are offsetand shown in alternating
We have corrected for the effects of air mass for all light curves and have fited a linear trend in time to both full-transit observations (the bottom two lightcurves).
RESEA LETT
Extended Data Figure 2 | Lombâ&#x20AC;&#x201C; Scargle periodograms of HARPS radial-velocity measurements. a, Periodogram of our window function for our radial-velocity observations. We see large peaks atapproximately 18 days and its harmonics. We also see large peaks at around 40 days and atthe stellar rotational period of 131 days (red dashed line). The
line. b, Periodogram for our HARPS radial-velocity observations. We see substantial power near the 24.73712-day orbital period of LHS 1140b (black dashed line), as well as broad power atlonger periods associated with stellar rotation (131 days, red dashed line) and at shorter periods associated with the window function of our observations.
LETT RESEA
Extended Data Figure 3 | Radial-velocity (RV) observations of LHS 1140 phased to our best-fitting sinusoid for the stellar activity signal. The signal from LHS 1140b has notbeen removed. Grey datapoints are copies of the purple data points and the red line
best-fiting model. We find a radial-velocity amplitude of 4.26 Âą 0.60 m sâ&#x2C6;&#x2019;1 due to stellar activity coupled with rotation, comparable to the radialvelocity amplitude of LHS 1140b. RV = 0 corresponds to the radial velocity 3 of the hoststar.
RESEA LETT
Extended Data Figure 4 | Lombâ&#x20AC;&#x201C; Scargle periodograms of HARPS radial velocity measurements. a, Periodogram of our residual radial velocities after subtracting our best-fiting sinusoid for the stellar activity signal. The highestpeak in this datasetis located atthe orbital period of LHS 1140b (black dashed line) and the broad power located atlong
been suppressed. b, Periodogram of our residual radial velocities after subtracting our best-fiting model thatincludes stellar radial-velocity variation as well as the orbitof LHS 1140b. We see no additional substantial peaks in our radial velocities, with the small peak located near P = 18 days due to the window function of our observations.
LETT RESEA
Extended Data Figure 5 | HARPS radial-velocity (∆RV) measurements fitted with a Gaussian process model. Our radialvelocity observations (points; error bars are 1σ) and best-fiting Gaussianprocess-based model (line with shaded 1σ error regions). The high cadence and adequate observational strategy allows us to identify the orbital signature of
incorporates the uncertainty thatarises from the activity-driven signals of the hoststar in our final determination of planetary mass. The residuals obtained after subtracting the model from the data are shown for the bottom two panels. ∆RV = 0 corresponds to the radial velocity of the hoststar.
RESEA LETT
Extended Data Figure 6 | Marginalized posterior distributions of the radial-velocity model parameters. The solid lines overplotted on the histograms are kernel density estimations of the marginal distributions. These smooth, Gaussian-shaped posterior distributions indicate the good convergence of the MCMC chain. Here we show the parameters of the Gaussian process (A, E, R and S), the orbital period of LHS 1140b (Pb), its radial velocity (relative to the hoststar; RV0),
semi-amplitude (Kb) and time of mid-transit(t0,b), the parameters and ecos() where eb is the eccentricity of LHS 1140b and esin()ω bbω bb, ωb is its argumentof periastron, and a white-noise error term (σs; in units of metres per second) thatis added in quadrature to each HARPS radialvelocity data point. The top leftis histogram is for A (the amplitude of the Gaussian process).
4
LETT RESEA
Extended Data Figure 7 | Near-infrared spectrum of LHS 1140. IRTF/SPeX spectrum of LHS 1140 (black), with M4V (blue) and M5V (red) spectraover-plotted. We classify LHS 1140 as an M4.5V star on the basis on its near-infrared spectrum.
RESEA LETT
Extended Data Figure 8 | Photometric modulation from the rotation of LHS 1140. The plotshows MEarth-South photometry of LHS 1140. We find thatLHS 1140 has photometric modulation due to stellar rotation and the asymmetric distribution of starspots, with a period of
from the two telescopes monitoring LHS 1140 are coloured in red and blue, and a sinusoid with a 131-day period is over-plotted in black. We find thatthe amplitude of variation increased between 2014 and 2016. Error bars, 1Ď&#x192;.