Tag Archives: Kepler mission

Alien megastructures

For the past five decades the search for extraterrestrial intelligence has been dominated by the search for radio signals. There are good reasons why this search strategy makes sense, but the available search space is so vast (our dishes have to be pointing in the right direction at the right time, and tuned to the right frequency) that the phrase “radio SETI” is an excellent synonym for the phrase “looking for a needle in a haystack”. Are there any other options for the search?

Personally, I believe that we need to adopt a Stapledonian approach to the problem.

Olaf Stapledon, a British philosopher and science fiction author, considered what might happen to intelligence in the distant future. For example in one of his novels, Star Maker, published in 1937, he described what we now call Dyson spheres: structures that orbit a star and enable a civilisation to utilise most of the energy output of its parent star. The creation of a Dyson sphere is far beyond our present technical capabilities, but who knows what we (or our mind-children) will be able to achieve a thousand years from now, or ten thousand years from now, or a hundred thousand years from now. And 100,000 years is an eye blink in cosmic terms; if there are extraterrestrial intelligences out there then they might be millions of years in advance of us. Thus in a Stapledonian approach to SETI we would look for examples of astroengineering, or megastructures that could only have been developed by technologically advanced intelligences. The detection of such a megastructure wouldn’t open up the possibility of communication, as a traditional radio SETI detection might do, but it would at least tell us that we were not alone. That in itself would be a terrifically important discovery.

The recent furore surrounding KIC 8462852 is an example of how a Stapledonian approach is starting to appear in the ongoing search for extraterrestrial intelligence. KIC 8462852 is an F-type main-sequence star about 1480 light years away from Earth in the constellation Cygnus. The Kepler space telescope recorded fluctuations in the light from the star – but a recent paper demonstrated that the fluctuations were so bizarre that they could not come from a transiting exoplanet. For one thing, the dimming of the starlight is not periodic; for another, the dimming is extreme (15% in one episode, 22% in another; for comparison, a Jupiter-sized planet blocks about 1% of its star’s light). So what is causing this weird behaviour? We don’t know. The authors of the paper suggest that the dimming might be caused by a series of comets, surrounded by clouds of gas, perturbed from their usual orbits by the gravitational influence of a nearby star; a small red dwarf close to KIC 8462852 might be the culprit. It’s possible. But it’s far from certain that this story can explain all the features that are seen. Any other explanations? Well … could it be that we have caught an advanced alien civilisation in the act of building a Dyson sphere? I doubt it. I REALLY doubt it. Just because we observe something we can’t immediately explain we shouldn’t immediately attribute it to alien intelligence (remember that, for a short while, the radio signals from pulsars were thought to be evidence for ETI; astronomers soon figured out the true explanation for the radio emissions). Nevertheless, it surely can’t harm to follow up these observations of KIC 8462852 with traditional radio-based SETI observations.

Several astronomers have already searched for the infrared emission that would accompany a Dyson sphere. But the search for Dyson spheres would form only a small part of a Stapledonian approach to SETI. We need to use our imagination and try to envisage the sorts of technology that a truly advanced civilisation might develop. In a previous post I looked at how John Smart’s transcension hypothesis argues that black holes are an attractor for intelligence. The philosopher Clément Vidal adopts a related approach. (Incidentally, in my book I wrote that is Belgian. He works in Belgium, but is in fact French. Apologies, Clément!)

Clément uses a two-dimensional metric, first proposed by John Barrow, to describe advanced civilisations. The Kardashev metric is well known: K1 civilisations control the energy output of their home planet, K2 civilisations control the energy output of their home star, K3 civilisations control the energy output of their home galaxy. But Barrow pointed out that there is a scale of inward manipulation that might be just as applicable to extraterrestrial civilisations: a B1 civilisation can manipulate the universe at the 1m level; a B2 civilisation can work at the 10–7m scale; a B3 civilisation manipulates at the nanoscale; and a BΩ civilisation can manipulate spacetime at the Planck level. In a 2011 paper Clément talks about the possibility of K2-BΩ civilisations; he has since switched to a more memorable appellation “stellivore”.

If we accept that stellivores could exist, an obvious question is: what might stellivores be doing that our telescopes and instruments might pick up?

Well, a stellivore might possess a technology involving black holes. (I won’t go here into the many reasons they might want to use black holes. Suffice it to say that a Stapledonian mindset would consider black holes to be a natural endpoint for many technologies.) And we know that in principle it is possible to detect activity around black holes; we know this because astronomers have already investigated X-ray binaries (XRBs). In an XRB a donor object (typically a star) loses material to a compact accretor (typically a black hole). The infalling matter releases huge amounts of gravitational potential energy. So could XRBs provide evidence for stellivores? In my book I write that we could “look for evidence for the regulation of energy flow within XRBs”. As Clément points out, there’s already evidence for such regulation; the key question – just as it is with KIC 8462852 – is whether the observations are best interpreted in terms of astrophysics or astrobiology?

Since Clément’s 2011 paper he has extended his vision to include a wider range of XRBs: the stellivore family could include cataclysmic variable X-ray pulsars, for example, with black holes being the end stage.

To my mind, the great thing about this Stapledonian sort of approach is that it broadens the range of techniques we can apply when searching for signs of extraterrestrial intelligence. Traditional radio-based SETI has its place. But the ideas of John Smart and Clément Vidal tell us that we could also profitably search at the highest energies.

A multiplicity of worlds

The day after my birthday the Kepler team announced (I like to think perhaps as a belated present) the discovery of 715 new exoplanets. This is a huge haul. It brings the tally of known planets to almost 1700.

The team was able to confirm the existence of such a large number of planets by making use of a new statistical approach to their analysis. Kepler worked by observing 160,000 stars and looking for periodic dips in brightness. The idea was that these periodic dips could be a sign of a transiting planet. The trouble is, these dips could also be caused by orbiting binary stars eclipsing each other. With the new technique, the Kepler team looked for multiple dips in brightness: this phenomenon must be caused by transiting planets rather than multiple eclipsing stars.

The technique works beautifully: those 715 planets just announced orbit only 305 stars. The Kepler data contains information on planetary systems, not just single planets.

Kepler tells us that planetary systems are common. It tells us small planets are common. And it tells us that some planets will orbit in the habitable zone. Deep-down we knew all those things anyway; but because of this announcement we can be sure.



Smaller than Mercury

That astronomers can find exoplanets at all is still a source of wonder to me. That they can find Earth-sized planets is astonishing. But a paper published in today’s issue of Nature is almost miraculous: A sub-Mercury-sized exoplanet, by Thomas Barclay and many others, describes the discovery of an exoplanet that has a radius that’s just 0.3 times that of Earth. It’s smaller than Mercury, in other words.

Kepler 37b, as it’s name implies, was found from data taken by the Kepler mission. The parent star, Kepler-37, is interesting because it’s the densest star in which solar-like oscillations have been detected. Just as a measurement of the frequencies of a musical instrument allows you to determine some of the properties of that instrument, the characteristic “ringing” of a star allows astronomers to determine some of the star’s properties with great accuracy. In this case astronomers were able to determine the radius of Kepler-37 with great precision, and this in turn allowed them to determine the radius of its planets with precision. Transit signals suggest that Kepler-37 has three planets. Kepler 37d has a radius about 1.99 times that of Earth’s; Kepler 37c has a radius about 0.74 times that of Earth’s; and Kepler 37b has a radius just 0.3 times that of Earth’s. It’s not much bigger than our Moon – and Kepler detected it!

Artist's impression of Kepler 37b

An artist’s impression of Kepler 37b (Credit: NASA)

For every 200 stars that Kepler studies you’d expect to see the transit signal in the data of perhaps one star. So the fact that the astronomers were able to identify this sub-Mercury-sized object does rather tend to suggest that small planets are extremely common.

Earths galore

The Kepler mission team announced exciting new results on 7 January 2013, at the 221st meeting of the American Astronomical Society. As I’ve explained in other posts (such as this one here), the team aims to detect exoplanets by staring at more than 150,000 stars in a fixed part of the sky. The Kepler telescope looks for regular dips in the brightness of these stars, variations that might be caused by the presence of a transiting planet.

The technique used by the Kepler team has been tremendously successful. On 7 January Christopher Burke, a Kepler scientist at the SETI Institute, announced the discovery of 461 new planetary candidates. So, as of the time of writing, Kepler has found 2740 potential planets orbiting 2036 stars. This is a hugely impressive number, when you consider that it wasn’t such a long time ago that people were debating whether exoplanets existed at all and, if they did, whether it would be possible to detect them.

The sizes of the planetary candidates discovered by Kepler(Credit: NASA/Kepler)

The sizes of the planetary candidates discovered by Kepler
(Credit: NASA/Kepler)

The AAS presentation that really caught the attention of the world’s press, however, was that given by Francois Fressin of the Harvard-Smithsonian Center for Astrophysics. Fressin has tried to estimate the fraction of stars that possess of Earth-sized planets, based on the Kepler data. Kepler will inevitably see only a small fraction of exoplanets because the transit technique only picks up those planetary systems in which the orbital plane is more or less side on to our view. If the orbital plane is slanted by more than a few degrees to our line of sight, the planets we won’t see the planets transit and there’ll be no drop in brightness. On the other hand, Kepler is seeing regular brightness fluctuations that are not due to transits; for example, a non-variable star might be extremely close in our line of sight to a regular variable, which would give a similar signal to a transit. After Fressin had corrected for both these effects he arrived at the following estimate: one in six stars host an Earth-sized planet. If that is indeed the case, there are about 17 billion Earth-sized planets in our Galaxy and it’s certain that some of these will be in the habitable zone.

If we’re looking to explain the Fermi paradox, we now know that we can’t invoke a paucity of Earths!

CHEOPS go-ahead

I’m old enough to remember a time when some people thought we’d never discover planets beyond the solar system. It’s not that long ago that astronomers first managed to confirm the existence of a few exoplanets. Techniques improved, and the number of known exoplanets started to increase rather rapidly. Then the Kepler mission was launched, and the number of candidate exoplanets simply rocketed (at the time of writing, Kepler has identified 2321 exoplanetary candidates). That’s impressive progress.

Now that astronomers have identified so many exoplanets, it becomes possible to design a mission that can study those bodies in more detail. That’s precisely what the CHEOPS mission will do. (Yes, CHEOPS is yet another acronym. This one stands for CHaracterising ExOPlanets Satellite.) ESA have selected CHEOPS for study as the first “small” or S-class mission.

If all goes well, CHEOPS will launch in 2017 and study stars that we know have planets around them. The satellite will monitor a star’s brightness, looking for the characteristic dip in brightness as a planet transits. This measurement will allow astronomers to determine the radius of the transiting planet; if the planet’s mass is already known from other measurements then astronomers will be able to calculate the planet’s density. This in turn will provide clues about the planet’s internal structure. CHEOPS will tell us a lot about the formation and evolution of planets with a similar mass to Earth. And, just as Kepler has provided targets for CHEOPS to study, CHEOPS in turn will provide targets for follow-up study by the next generation of powerful telescopes such as E-ELT.

An artist's impression of CHEOPS

An artist’s impression of CHEOPS. The satellite will be placed in a Low Earth Sun-Synchronous orbit at an altitude of 800km. Its 33cm telescope will observe in the range 400-1100nm.
(Credit: University of Bern)


Solution 42 in my book Where is Everybody? is entitled “The Moon is Unique”. What has the Moon got to do with the Fermi paradox? Well, it seems quite likely that our Moon has played an important role in the development of life on Earth (for example, it stabilises Earth’s axial tilt and thus prevents extreme climatic variations) and it’s not entirely implausible that it played a role in the creation of life in the first place. However, the Moon was created in a giant collision between Earth and a Mars-like object. Had the parameters of that collision been slightly different, our Moon would not have formed with the size it has – and its effect on life would have been different. So, the argument goes, an Earth-Moon system such as our own might be rare – and so therefore might life.

I don’t believe that a scarcity of moons resolves the Fermi paradox – but for all sorts of reasons it would be good to understand more about moons in other planetary systems. And large moons – satellites such as Saturn’s Titan, for example – could themselves be hosts for life. The difficulty, of course, is in finding exomoons.

A recent paper by Kipping, Bakos, Buchhave, Nesvorny and Schmitt (The hunt for exomoons with Kepler (HEK): I. Description of a new observational project) explains how we might be able to search for exomoons with current technology.

How Kepler searches for exoplanets

Kepler searches for exoplanets by looking for a periodic dimming caused by a planet transitting a star
(Credit: NASA)

The Kepler mission, as we know, searches for exoplanets by looking for the periodic dip in a star’s brightness that occurs when a planet transits the star. The technique has resulted in the discovery of hundreds of planets. Well, suppose the planet has a moon that orbits in more or less in the same plane as the planet orbits its star: when planet and moon were side-by-side they would block more light than when one object was in front of another. By looking in detail at the periodic variation in brightness of the star it should be possible, in principle, to determine the moon’s mass and diameter (and hence its density).

In the paper, Kipping and his co-authors calculate that Kepler should in principle be able to discover exomoons with a mass as small as 0.1 Earth masses. Such an object would be much bigger than Ganymede or Titan. The discovery of such an exomoon would be important for science, but would not in itself shed much light on the question of extraterrestrial life (other than increasing, perhaps, the potential number of abodes for life). But if the hunt for exoplanets has taught us one thing, it’s that observations that once appeared technically impossible eventually become routine. Right now it might be impossible to search for an exomoon that’s similar to our own Moon. In a few years time it won’t be.

Citizen science for astrobiology

I was one of the early adopters of SETI@Home – the pioneering distributed computing project that uses a volunteer’s spare computing capacity to crunch data from Arecibo in the hope of finding extraterrestrial signals. I donated my spare CPU cycles as long ago as 1999.

SETI@Home has been hugely influential, of course, but at heart it’s a passive activity. The same comment applies to many of the projects it spawned, such as Einstein@Home. These are all valuable endeavours, but they don’t require much from a volunteer except that donation of computing power.

The newer “citizen science” projects exemplified by Zooniverse require active involvement from volunteers: these projects succeed, and they produce real scientific results, because the human brain is still better than a computer program at pattern-recognition tasks. Zooniverse contains one project that is clearly of interest to astrobiology: planethunters.org gets volunteers to search for explanets using lightcurve data from Kepler. And other projects in the Zooniverse stable might well turn up some serendipitous discoveries: who knows, for example, what people might see when they look through the hundreds of thousands of images in Galaxy Zoo?

But the prime focus of these projects is not astrobiology. Are there any potential “citizen science” projects directly relating to astrobiology? Not just the passive donation of spare CPU cycles in a search for extraterrestrial signals, but the active involvement of hundreds of thousands of human brains on astrobiological problems. Does anyone have any ideas what such a research program would look like?

Earths are common, not rare

The paper One or more bound planets per Milky Way star from microlensing observations, which appears in today’s Nature, settles the question of whether planetary systems are common.

The paper’s 42 authors (who constitute a truly global collaboration of astronomers) analysed gravitational microlensing data gathered in the period 2002 to 2007. Now, gravitational microlensing – the short-term brightening of a star that occurs when the star, the planet of an intervening star and one of our telescopes move into alignment – is rare: at any particular time, less than one star in a million will be subject to a microlensing event. So although gravitational microlensing is a well-established planetary detection technique, it is less productive than either the radial velocity method (which is used by experiments such as HARPS) or the transit photometry method (which is being used by the phenomenally successful Kepler Mission). Thus the present paper does not announce the sudden discovery of a large number of previously unknown exoplanets. Rather, by statistically analysing the number of events that were detected, the team was able to estimate how many exoplanets are likely to exist out there.

The bottom line is: stars are more likely than not to possess planets. Our Galaxy must be teeming with planets.

Drake equation

The Drake equation
Credit: Waifer X (Flickr CC)

The Drake equation contains a term fp – the fraction of stars formed that will have planets. Once upon a time, not that long ago, some astronomers believed that planets were rare. Perhaps, they argued, fp was small. Now we know for certain that isn’t the case. We know for certain that we can’t look at this term as a solution to the Fermi paradox.

It seems to me that every astronomical advance is simply sharpening the paradox.

Water on 22b, or no water on 22b? That is the question.

NASA’s Kepler mission is proving to be a huge success, better than I thought was possible. Prior to Kepler, most planets were found using the Doppler spectroscopy method, in which telescopes look for the tell-tale to-and-fro movement of a star’s spectral lines as the star itself wobbles due to the gravitational tug of a planet. Kepler uses a different technique. It looks for the tiny dip in a star’s brightness when a planet transits – it’s the equivalent of looking for the dimming of a car’s headlight when an insect flies across it.

Artist's impression of Corot-7b transitting the yellow dwarf Corot-7

An artist's impression of Corot-7b transitting the yellow dwarf Corot-7: note how small the planet is in relation to its parent star
Credit: Kevin Heider

For any given star it’s unlikely that we’ll see a transit. But Kepler observes the brightness of so many stars at once – there are about 145,000 main sequence stars in its field of view – that the numbers stack up in its favour. Thanks to Kepler, the transit photometry method of exoplanetary discovery is proving to be much more productive than the Doppler method: at the time of writing it has found 2326 candidate planets. (Progress in this field never ceases to amaze me. When I was a student it seemed it impossible to detect exoplanets on this scale.)

The Kepler discovery announced by NASA on 5 December 2011 is perhaps the most interesting yet. The planet Kepler- 22b is in the habitable zone: in other words it’s possible that liquid water flows on this planet. We shouldn’t get carried away after this announcement: this planet is much bigger than Earth (about 2.4 times Earth size) and we don’t know whether it has an atmosphere (and the presence of an atmosphere will strongly influence the planet’s surface temperature). Nevertheless 22b is the most Earth-like planet we’ve ever seen. It orbits a star that’s like our Sun, it orbits in the Goldilocks zone, and it even has a year (290 days) that’s similar to Earth’s year.

Artist's impression of Kepler-22b

An artist's impression of Kepler-22b
Credit: NASA

The big question is: does 22b, like Earth, possess liquid water? And that’s important, of course, because ‘life as we know it’ requires water.

Even if it turns out that 22b isn’t an Earth-twin – and my guess is that in the end it will turn out to be not much like Earth at all – Kepler will eventually find a planet that’s truly Earth-like. And even if Kepler doesn’t, then some other mission will do so. (I’m not sure what that mission will be: the European Darwin mission has been cancelled, as have the American Terrestrial Planet Finder and SIM Lite missions. But some experiment, some time, will find an Earth twin.) I’m certain that there are myriads of truly Earth-like planets out there, all of them potentially homes for life. To my mind, then, the really interesting thing about planets like Kepler-22b is how it makes Fermi’s question even starker:

Where is everyone?