WEBVTT 00:00.240 --> 00:04.720 From Dyson swarm to comet apocalypse—it's  been 10 years since the discovery of Tabby’s   00:04.720 --> 00:10.748 star and we finally have a leading contender to  explaining the strangest known star in the galaxy. 00:16.368 --> 00:23.648 The galaxy swarms with worlds. Almost every one  of its hundreds of billions of stars boasts a   00:23.648 --> 00:29.488 system of multiple planets. Some may be much  like the Earth. We know this because the Kepler   00:29.488 --> 00:35.168 mission has clocked the miniscule but extremely  regular darkening of thousands of stars due   00:35.168 --> 00:41.328 to their planets passing in front of them. It’s the regularity of these transits that   00:41.328 --> 00:47.008 tells us that the shadow is caused  by the repeated orbit of a planet.  00:47.008 --> 00:53.728 So what do we make of a light curve like t his? You’ve probably seen this before—it’s the light   00:53.728 --> 00:59.168 curve of KIC 8462852—better known as  Tabby's Star after its discoverer,   00:59.168 --> 01:05.728 Tabitha Boyajian. These dips are not due  to planets. They’re not regular in either   01:05.728 --> 01:11.328 repetition or depth. Sometimes up to 20%  of the star’s light is somehow blackened.  01:11.328 --> 01:15.648 For a while, Tabby's Star became one of  astronomy's greatest mysteries. What could   01:15.648 --> 01:21.568 possibly be causing these shadows? It inspired  explanations ranging from swarms of comets to   01:21.568 --> 01:26.928 giant ring systems—and even a brief fascination  with the possibility of alien engineering.  01:26.928 --> 01:33.088 The galaxy’s strangest star dropped out of th e  headlines in the decade since, but astronomers   01:33.088 --> 01:39.488 continued to study it intensely. Observation after  observation, some hypotheses were ruled out and   01:39.488 --> 01:45.968 others bolstered. And just last month a new paper  came out that may have found a hidden object in   01:45.968 --> 01:52.928 orbit around Tabby’s star. Not the thing making  the shadows. The thing making the shadow-makers.  01:52.928 --> 01:56.608 The Kepler mission worked so well because  planets are supremely predictable,   01:56.608 --> 02:03.168 and so are the shadows they cast. They’re very  round objects following pretty round orbits.   02:03.168 --> 02:08.688 We see tiny but consistent dips in their host   star’s light exactly once per orbit when they   02:08.688 --> 02:14.768 pass in front — for the small fraction of stars  whose planetary systems happen to be alined to   02:14.768 --> 02:21.008 allow such a transit from our perspective. The  time between dips gives the orbital period,   02:21.008 --> 02:26.128 the duration of the dip gives the speed of motion,  the depth of the dip and the duration of the dip’s   02:26.128 --> 02:31.648 onset gives the size of the planet; and together  this is enough to pretty thoroughly characterize   02:31.648 --> 02:38.928 the world. All that from its shadows—its very  clean and distinctive shadows. But shadows over   02:38.928 --> 02:45.248 Tabby’s star are far from clean and distinctive,  and are certainly not from simple planets.  02:45.248 --> 02:51.248 Now Kepler also worked really well because  it’s supremely specialized. A camera designed   02:51.248 --> 02:57.008 to detect the miniscule dimming of stars out to a  few thousand light years, with a field of view to   02:57.008 --> 03:04.368 monitor over 150,000 stars over its 5 year primary  mission. But specialization comes at a cost—and   03:04.368 --> 03:10.048 one of the costs was that its pictures  are all black and white. More technically,   03:10.048 --> 03:15.168 it had a single filter and so had no way  to collect spectral or colour information.  03:15.168 --> 03:19.728 That’s OK, because when a planet passes in  front of a star, it blocks all light equally   03:19.728 --> 03:24.928 and spectral information isn’t that useful.  A planet is as opaque to blue light as it is   03:24.928 --> 03:32.368 to red light. We call such colour-independent  dimming “achromatic” or “grey”. The same should   03:32.368 --> 03:37.728 be true of any solid, opaque object causing the  shadow. That includes a planetary ring system,   03:37.728 --> 03:42.528 whose icy particles are large enough to block  all light they intercept, and certainly most   03:42.528 --> 03:48.368 technological structures that we can imagine. But not everything casts a grey shadow. Some   03:48.368 --> 03:54.848 “absorption” is chromatic—the darkening depends on  the wavelength of the light. For example, Earth’s   03:54.848 --> 04:00.448 atmosphere blocks short wavelength light more  easily—blue more easily than green more easily   04:00.448 --> 04:05.728 than red. That scattered blue light gives us our  blue daytime sky, and the increased path through   04:05.728 --> 04:11.808 the atmosphere during sunset increasingly saps  shorter wavelengths, reddening the setting sun.  04:11.808 --> 04:17.648 We also see this effect out there in the universe.  The most common type of chromatic absorption is   04:17.648 --> 04:22.848 due to tiny particulates—heavy elements blasted  out by dying stars that have congealed into tiny   04:22.848 --> 04:29.808 grains. In other words, by dust. Obscuration  by dust was another leading contender for the   04:29.808 --> 04:36.608 shadows over Tabby’s star, and here at last we  have a real way to test a hypothesis. Planets,   04:36.608 --> 04:41.088 giant ring systems, Dyson swarms,  should produce gray absorption. Dust   04:41.088 --> 04:46.128 should produce chromatic absorption . In 2018 the Boyajian team published new   04:46.128 --> 04:50.048 observations of the star, in which they showed  that the dimming is indeed stronger at shorter   04:50.048 --> 04:54.528 wavelengths, just as you’d expect from giant  clouds of dust passing in front of the star.  04:54.528 --> 05:00.688 Yep, dust. After years of speculation, you’d  be forgiven if you were disappointed at the   05:00.688 --> 05:05.968 apparent ordinaryness of this explanation.  But there’s a reason we never, ever see   05:05.968 --> 05:11.808 dust absorption around stars like Tabby’s. In  fact, astronomers were as perplexed as ever.  05:11.808 --> 05:17.888 That’s because dust can never just sit there  near an active star—the pressure of the stellar   05:17.888 --> 05:24.768 radiation will quickly sweep it away or destroy  the tiny grains. Now we do see thick dust around   05:24.768 --> 05:29.808 some stars. Very young stars are often surrounded  by leftover material from their formation. But   05:29.808 --> 05:35.488 that dust is quickly blasted away early in the  star’s life. And very old stars can eject dust in   05:35.488 --> 05:40.928 their final phases. But that dust only persists  because it’s constantly being replenished.  05:40.928 --> 05:45.808 Tabby’s star is neither young nor old. It’s a  regular main-sequence star about 40% more massive   05:45.808 --> 05:51.088 than the Sun, and is in the prime of its life. The  leftover dust of its birth should be long gone,   05:51.088 --> 05:55.888 and it isn’t producing dust itself. And  yet it's been showing these strange,   05:55.888 --> 06:02.128 varying dust shadows for years. So where  does the stuff come from? That’s now the   06:02.128 --> 06:09.248 mystery. The dust must be replenished  from somewhere. But from where, and how?  06:09.248 --> 06:16.608 Let’s return to good ol’ Sol for some inspiration.  Our solar system does have a source of new dust.   06:16.608 --> 06:22.928 From comets. Way out beyond Neptune are debris  fields that stretch a third of the way to the   06:22.928 --> 06:27.648 nearest stars. There, frozen chunks of material  leftover from the Sun’s formation drift through   06:27.648 --> 06:34.448 the darkness on vast orbits that can take millions  of years. But, now and then, one of these objects   06:34.448 --> 06:40.528 will find its way to the inner solar system.  There it becomes a comet. It begins to melt in the   06:40.528 --> 06:46.208 increasing solar radiation and sheds a long tail  of water vapor and … guess what … dust. When it   06:46.208 --> 06:51.888 was trapped in ice, that dust had been protected  from radiation, but once freed it forms a streak   06:51.888 --> 06:57.808 of material—the cometary tail—that persists  only until it’s dispersed by solar radiation.  06:57.808 --> 07:02.528 This happens only rarely. Objects out in the  vast Oort cloud can be gravitationally nudged   07:02.528 --> 07:07.088 by passing stars to fall into the solar  system. Those in the more nearby Kuiper   07:07.088 --> 07:11.648 belt are slowly nudged until they have close  encounters with gas giants that can smack them   07:11.648 --> 07:16.928 further inwards. On their new elliptical orbits,  these comets zip through the inner solar system   07:16.928 --> 07:23.168 and then spend between decades and gigayears  in the outer reaches. It’s very rare to find   07:23.168 --> 07:29.088 more than one comet in the inner solar system. So if we want to explain Tabby’s star’s situation,   07:29.088 --> 07:36.208 we need a lot more comets than this. Swarms of  them. We need some mechanism that’s constantly   07:36.208 --> 07:42.768 flinging comets at the star. Now maybe  alien spacecraft are peppering the inner   07:42.768 --> 07:47.888 planets to soften them up for an invasion, or  maybe to seed the barren planets with enough   07:47.888 --> 07:54.608 water to begin terraforming. But—much more  likely— is there's a massive, invisible body   07:54.608 --> 08:00.528 that’s causing gravitational disruption on a  much larger scale than we see in our system.  08:00.528 --> 08:04.608 This idea of a hidden gravitational  disruptor is reasonably convincing,   08:04.608 --> 08:10.608 and the idea has been around for a while, however  a dark mass orbiting a distant star is next to   08:10.608 --> 08:15.808 impossible to see. So how do we test this? In 2019, the Transiting Exoplanet Survey   08:15.808 --> 08:21.248 Satellite—TESS__watched Tabby’s star for 55 days  to study the strange, irregular dips. It happened   08:21.248 --> 08:27.248 to record a single, isolated, and unusually  symmetric dip during that period. At the time,   08:27.248 --> 08:31.728 it was thought to be a slightly odd dust  dip. But over the following few years,   08:31.728 --> 08:36.928 researchers failed to produce convincing models  that could explain the dip in terms of dust.  08:36.928 --> 08:41.488 Then, just this year, a team of researchers  led by Cristina Madurga-Favieres put forward   08:41.488 --> 08:48.208 a new proposal. Perhaps this dip is not dust but,  rather, the dust maker. Perhaps it was exactly   08:48.208 --> 08:54.528 the type of planetary transit event that Kepler  and TESS were built to discover—in this case,   08:54.528 --> 08:59.568 the massive body hypothesized to  be flinging comets at Tabby’s star.  08:59.568 --> 09:04.448 Let’s take a closer look at the intriguing  shadow. It blocks only about one percent of   09:04.448 --> 09:09.648 the star’s light, is just over 20 hours  long, and seems highly symmetric. This   09:09.648 --> 09:16.288 is what planetary transits look like. In fact,  had Kepler first seen a simple dip like this,   09:16.288 --> 09:21.808 rather than these crazy ones, researchers would  have immediately guessed “planet”. It was only   09:21.808 --> 09:26.368 in the context of the strange dimming that this  wasn’t immediately flagged as a likely planet.  09:26.368 --> 09:30.368 But when we learned that the crazy  dimming is probably due to dust,   09:30.368 --> 09:37.568 the existance of a gravitational disruptor to rain  comets down on Tabby’s star became a prediction.   09:37.568 --> 09:42.608 Which is why this dip g ot a second look. The challenge now is that we can’t confirm   09:42.608 --> 09:47.648 that a planet caused this shadow. At least  not yet. Normally, to confirm a planet via   09:47.648 --> 09:54.448 the transit method we need to see three transits.  Two full revolutions since the first one. If those   09:54.448 --> 09:58.608 two transit separations take the same amount of  time, then it’s a smoking gun for an orbiting   09:58.608 --> 10:06.288 planet. Here we only have one transit, so it’s  only tentatively a planet until two more cycles.   10:06.288 --> 10:13.088 But not content with just waiting—potentially  years—to see if the dip repeated, the team looked   10:13.088 --> 10:18.928 backwards in time at old data from Kepler, TESS,  and a suite of other monitoring programs for signs   10:18.928 --> 10:25.248 of past dips. They found none consistent with  a planet, but because of that they were able   10:25.248 --> 10:31.968 to rule out many possible orbits. In fact they  were left with a few orbital periods clustered   10:31.968 --> 10:38.368 between 1000 and 1300 days as being consistent  with no past transit observation. Factoring in   10:38.368 --> 10:44.448 the 20-ish day duration of the transit to get  the actual speed, this gives an orbital period   10:44.448 --> 10:50.848 of 3.3 years and an orbital radius of something  like 2.5 times Earth’s—placing it between Mars   10:50.848 --> 10:59.168 and Jupiter in our solar system. From the 1.1%  dimming it causes, it should be around 70% larger   10:59.168 --> 11:05.408 than Jupiter in size. That’s a huge planet. Earlier we were talking about comets coming   11:05.408 --> 11:10.928 from the Kuiper belt and Oort cloud. This star  is 43% more massive than the Sun and probably   11:10.928 --> 11:16.768 has a commensurately larger Oort cloud. Too big  to feel the gravity of a planet that’s so close   11:16.768 --> 11:22.128 to the star. However, it’s also possible that  this system has a much tighter ring of icy debris   11:22.128 --> 11:28.128 similar to our Kuiper belt. If this super-Jupiter  is massive enough, perhaps it disrupted outer   11:28.128 --> 11:34.128 planet formation leaving a rich field of  debris for it to throw at the inner system.  11:34.128 --> 11:38.048 But for that it would want to be very  massive. And as it happens, the next   11:38.048 --> 11:44.608 stage of this analysis can not only verify the  planet’s presence, but also measure its mass.  11:44.608 --> 11:50.448 After transits, the next most productive way to  find planets is the radial velocity method. A   11:50.448 --> 11:56.128 planet doesn't actually orbit its star—it orbits  its mutual center of mass with its star. And   11:56.128 --> 12:02.528 that star orbits that same mutual center of mass.  Because stars are much more massive than planets,   12:02.528 --> 12:07.008 this mutual orbital point is much, much  closer to the star—often actually inside   12:07.008 --> 12:13.328 of it. But stars do still move—or at least  wobble—in little circles due to each of their   12:13.328 --> 12:18.608 planets. That motion can be detected by  watching for the tiny shift in wavelength   12:18.608 --> 12:25.248 of the star’s light due to the doppler effect. Now the team used both old and new spectroscopic   12:25.248 --> 12:29.728 data to find evidence of that wobble. It’s  especially difficult for Tabby’s star because   12:29.728 --> 12:35.968 the star itself is rapidly rotating, which causes  additional Doppler smearing of its light. However,   12:35.968 --> 12:41.648 the team were able to find a tentative Doppler  signature that’s consistent with the orbital   12:41.648 --> 12:49.168 parameters they’d found. And the mod  el that best fit the radial velocity   12:49.168 --> 12:53.728 analysis gives , a ridiculously massive  planet—10 times that of Jupiter. That’s   12:53.728 --> 12:59.008 encouraging if we want some gravitational  beast that’s able to disrupt outer orbits.  12:59.008 --> 13:04.048 The radial velocity detection is quite  tentative. The confidence is 2.3 sigma,   13:04.048 --> 13:08.128 which means there’s a 1 in 50 chance that  the radial velocity signal was a conspiracy   13:08.128 --> 13:14.128 of random noise. It doesn’t yet qualify as a  sure detection, and so more radial velocity   13:14.128 --> 13:19.648 data is needed. But we also need more independent  approaches—and we’re about to have the capability   13:19.648 --> 13:25.488 to measure this stellar wobble in a way that was never before possible. Come December this year,   13:25.488 --> 13:30.528 we’ll be able to measure the literal change in  position of stars as they move in tiny circles   13:30.528 --> 13:35.568 due to their planets. The Gaia spacecraft was  designed to make the most precise position   13:35.568 --> 13:41.328 measurements ever achieved. It operated from 2014  until March this year. And we’ve already learned   13:41.328 --> 13:46.528 an insane amount about the positions and motions  of Milky Way stars, but perhaps the most ambitious   13:46.528 --> 13:51.648 and most difficult measurement it will make is  of what we call the “astrometric wobble” of a   13:51.648 --> 13:57.408 star due to its planets. And in December, the Gaia  team will have completed the highest level of data   13:57.408 --> 14:03.408 processing for the first 5.5 years of the survey  and will release precision astrometry for that   14:03.408 --> 14:09.728 period. They expect to discover 1000s of exoplanet  systems this way. The Tabby’s star team claim   14:09.728 --> 14:17.088 that its astrometric wobble due to its tentative  super-Jupiter will be visible in this data. If so,   14:17.088 --> 14:23.328 the mystery of Tabby’s star will be largely put  to rest. The star is being eclipsed by evaporating   14:23.328 --> 14:29.808 swarms of comets, flung inwards by a colossally  large planet. And if the planet’s not there? Well,   14:29.808 --> 14:37.088 we keep on thinking, and watching, and wondering  about the strangest star in local spacetime. 14:29.808 --> 14:35.488 a time of massive geologic upheaval full of  incredible creatures, which all culminates in the   14:29.808 --> 14:38.608 greatest mass extinction in the planet’s history.  Check out the link in the description to watch.