Are we alone in the Universe? For years, people have been making predictions, many using theDrake equation. That involves the use of various educated guesses about the frequency of planets, how many are habitable, and so on. Until about a decade ago, most of the values in the equation remained just that, however: guesses.
In the last dozen years, we've witnessed an amazing transformation in science and appear to be on the verge of several more. The existence of planets orbiting other stars—exoplanets—has gone from a hypothetical to a reality. We've now got a catalog of thousands of potential planets. In many cases, we even have an idea about their size, composition, and temperature. Some of them orbit stars that are, in galactic terms, right next door.
The result has been an incredible buzz of information—over the course of this winter, there were a series of updated estimates on the number of planets in the galaxy (answer: lots) along with various ways of slicing and dicing the numbers. How many Earth-like planets? How many orbiting stars like our Sun? In every case, the numbers were staggeringly large, with the possibility Earth could be one of millions, if not billions, of similar planets in our galaxy alone.
Given all this information, it seems like we're on the verge of finding Earth's twin—a small, rocky, planet sitting at just the right distance from a star to play host to liquid water. But that poses a far more significant challenge than what might be apparent from the field's most recent successes. An understanding of the challenges involved suggests a different time frame, one where we might still be decades away from getting a clear picture of what our galaxy's planets look like.
Spotting exoplanets
There are two main methods we've used for identifying exoplanets: radial velocity and transit. A number of other planets have been spotted through other means, but these two account for the vast majority of sightings.
Radial velocity relies on the fact that gravitational attraction is a two-way street. A star may exert a massive pull on the planets that orbit it but, on a smaller level, those planets pull back. As they swing through their orbits, their gravitational pull tugs the star ever so slightly in the direction of the planet. This motion is enough to cause a slight Doppler shift, the compression or expansion of wavelengths caused as objects move towards or away from an observer. As a star is pulled towards Earth, all of its light gets ever slightly more blue; as it is pulled away, it gets a bit redder.
To track these changes, telescopes have to be fitted with a special instrument called a spectrometer, which tracks how much light is emitted at specific wavelengths. And you have to observe a specific star for an extended period of time to make sure you can spot an acceleration that won't even be visible during portions of its orbit. There are also long- and short-term variations in a star's output (one example is the equivalent of sunspots) that make identifying planets a challenge.
Still, the method is highly effective, and it was the first used to identify an exoplanet. It's especially good if the planet is massive or close to the host star (or both). As a result, many of the first planets we found were what are called "hot Jupiters," gas giants that orbit close in to their host star.
The WASP telescope, a ground-based planet hunter.
One nice aspect of the radial velocity method is that it provides some indication of the planet's mass, since that determines the strength of its pull on the star. It actually provides a lower limit, since the magnitude of the Doppler effect will be largest if the planet's orbital plane lines up with the line of sight from Earth. You could get the same effect with a heavier planet that has an orbital plane tilted away from Earth.
The alternative is the transit method, which watches for signs of a planet passing between its host star and the Earth. This creates a tell-tale dip in the output from the star, as the planet creates a mini-eclipse. This definitely requires the planet's orbital plane to be lined up so that it runs through Earth. Instead of mass, this method provides an indication of the planet's size, since that determines how much light it can block out.
Combining the two methods gets you both size and mass, which lets you calculate the density of the planet. In multi-planet systems, you can also get this from the transit alone, as the gravitational interactions among the planets can slightly change the timing and duration of the transits.
There are a few other methods used to spot planets, and we've been able to image a number of them directly. But, for the most part, these two account for most of the planets we've observed.
What have we seen?
Initially, there were no instruments designed to detect exoplanets. Instead, researchers had to point existing instruments at a star—and then continue doing so night after night in order to track transits or changes in the velocity of the star. With instrument time at a premium, it's hard to arrange that sort of schedule. So, as a result, the first exoplanets to be spotted were the easiest ones to spot: large, Jupiter class planets orbiting in near their stars. For the first few years of exoplanet hunting, these were the vast majority of the exoplanets spotted.
This raised an obvious question: were we seeing so many just because they were easy to spot, or did this population actually represent the majority of the planets out there? To answer that question, scientists began to design dedicated instruments specifically intended to spot more plants, someground-based, at least one based in space. In at least one case, a major instrument was fitted with a spectrograph that has specialized in planet hunting. Combined, these instruments began expanding our catalog of exoplanets.
The Kepler planet-hunting telescope.
And, in the process, they began to change our picture of our galaxies' planets. Smaller bodies—warm Neptunes and super-Earths—began to appear in the catalog. But we still didn't have a complete enough picture to start making inferences about what the galaxy as a whole looked like.
The Kepler mission was intended to change that, and it has succeeded spectacularly. The space telescope stares down one of the spiral arms of our galaxy, with nearly 150,000 stars in its field of view. Over the past several years, it has found the tell-tale dips in the light that indicates a planet is passing in front of one of them. So far, 105 of these have been confirmed to arise from a planet; there are another 2,740 candidates waiting to be confirmed. We can now start to do statistics.
One of the key things Kepler told us is that most planets are far smaller than Jupiter. First, it became clear there were a lot of Neptune-equivalents, and later, Earth-sized bodies started showing up in the data. It quickly became obvious the numbers went upwards as planet size went down. Rather than being filled with Jupiters, the majority of the planets in our galaxy look much more like Earth. With time, another trend became apparent: the numbers went up the further you got from the star. Not everything was likely to have molten metals bubbling on its surface.
Even two years ago, Kepler's planetary haul was heavy on Neptune-sized planets and super-Earths. Now, Earth sized planets are the most common things identified.
So, that tells us something about the typical planet. How typical are they? Quite. Detecting a planet using Kepler means the system's orbital plane must be edge-on when viewed from Earth. Given the probability of that happening (which is purely a matter of geometry), we can extrapolate out to how many planets must be in Kepler's field of view. And from that, we can estimate how many planets there are in the galaxy total.
The Milky Way contains about 300 billion stars. On average, each of them has a planet (although many of these are in multi-planet systems, meaning many stars have none). Our galaxy is teeming with planets.
Running the numbers
Kepler may have given us a picture of what our galaxy was like, but to the people who helped design and run it, the telescope was just the first step in a multi-decade plan. The goal is to identify a planet that is likely to host life. The basic idea is, using Kepler, you can get some sense of what the distribution of planets in the galaxy is (which we have). Then, using that, you could work backwards to get a sense of what the population of planets is likely to be in the area around the Solar System.
From that, you could infer how far away you'd have to look in order to have good odds of imaging multiple small, rocky planets that are likely to be in the habitable zone. And, using that information, you could determine the sort of hardware you'd have to build in order to have a good shot of imaging their atmospheres. It's possible the chemical composition of the atmosphere, and maybe light reflected off its surface, would give us some hint of whether there's life there.
So, with several years of Kepler data, how are things looking? Pretty good. One recent analysisfocused on small, cool stars called M dwarfs, fairly common in the neighborhood of the Sun. Based on the number of planets spotted orbiting these stars by Kepler, two researchers calculate there's a 95 percent chance of finding an Earth-sized planet orbiting in an M dwarf's habitable zone using the transit method. If we don't require it to transit (in other words, have the planet's orbital plane align with Earth), then the 95 percent probability zone drops to under 25 light years.
And that's just a single type of star; there are representatives from a number of other types within Earth's vicinity. Clearly, if we want to build the hardware to start looking at local planets, the statistics generated by Kepler suggest that it wouldn't have to be able to peer too deeply to spot some.
If we build it, what will we see?
But would they be habitable? That's actually a much harder question. As the paper linked just above put it, "The concept of a 'habitable zone' within which life could exist is fraught with complications due to the influence of the spectrum of the stellar flux and the composition of the planetary atmosphere on the equilibrium temperature of a planet as well as our complete lack of knowledge about alien forms of life." Let's look at each of these issues.
As far as life is concerned, we have just a single sample. It is mostly comprised of the elements that are common in the Universe, like hydrogen, carbon, nitrogen, and oxygen. Life as we know it is also water based. Are any or all of these requirements? It's impossible to know. Some people have considered the possibility of alternate solvents, like liquid ammonia; there's also the chance that life on a different planet relies on elements that, by chance, are abundant in that specific environment.
Still, because water has a number of characteristics that make it very appealing for complex chemistry, most concepts of a habitable zone have focused on finding conditions that can support liquid water—meaning a temperature between 0 and 100C. So, knowing how hot a given star is should tell us where that zone should be, right? Well, no, as the quote above should make clear.
Although you might not realize it from the news, the greenhouse effect actually exists, and can be caused by chemicals—water vapor, methane, carbon dioxide—that are likely to be common on other planets. So, one key feature that will determine whether a planet is habitable is the existence of an atmosphere. And that's not an easy thing to predict. The initial composition of a planet and its tectonic activity will determine what the raw materials are, but then these interact in complex ways with things like the planet's gravity and the presence of a magnetic field.
Closer to the host star, things are even more complex, as detailed in another recent paper. Any star prone to violent outbursts in the extreme UV would be able to dissociate things like water in the atmospheres of nearby planets, allowing the hydrogen to escape into space. And, if the atmosphere is diffuse enough, the outer layers of it will end up slowly drawn off by gravity, pulled into the host star. Despite all these hazards, the authors conclude atmospheres can survive on Neptune-sized planets, even if they are close to the star they orbit.
The color of feedbacks
Then there are issues of feedbacks. Let's say you have a water-rich planet. Too close to the host star, and the water will heat up enough to enter the stratosphere and escape. Too far away, and the water will condense out of the atmosphere. Water tends to absorb light well, heating things up, but it can also form ice, which generally reflects sunlight into space. Carbon dioxide works in a similar way: if a planet gets cold enough, it will freeze directly out of the atmosphere into a solid, creating a strong cooling feedback that will cool the planet further.
But these feedbacks also vary based on the type of star the planet is orbiting. For example, water vapor scatters blue light more efficiently, and will therefore have a stronger cooling effect near hot, blue stars. Water ice, in contrast, is actually capable of absorbing some wavelengths of infrared light. That could cut significantly into the cooling feedback normally provided by ice.
To give a better sense of how complicated it is to figure out where a habitable zone should reside, it's worth taking a look at this paper, which tries to model the atmospheres of exoplanets as a one-dimensional stack of atmospheric layers. It tries to account for all of these factors (and a whole lot more), but somehow comes up with the result that the habitable zone around the Sun ends one percent closer to the Sun than Earth's orbit.
Obviously, the Earth seems to be in little danger of having its oceans boil off, which provides some indication this is still an imperfect science. In the same way, the best models have trouble giving Mars the warm and watery past that all the geological evidence points to.
All in all, it seems that trace gasses and things like the planet's surface properties (how much light it absorbs and reflects) are likely to play key roles in determining a planet's average surface temperature.
Hidden in the averages
Of course, that's just the average temperature. As we see from Earth, our average is produced by a wide range of temperatures, some quite hospitable to life, others that would forbid it. And Earth's average has changed dramatically over time. It might not have been possible to form life (or at least our version of it) at a time like the snowball Earth, when there were glaciers in the tropics. But life had already diversified by then, and at least some organisms were able to adapt to these conditions (or at least some niches within them). So, even knowing a planet's current average doesn't necessarily rule out the presence of living creatures, though within limits—it's hard to imagine refuges for life in a Venus-like hell.
Researchers are also considering the possibility of life in habitats very much unlike Earth's. To give one example, planets orbiting a dim red star need to be quite close to it in order to have a chance of hosting water. At those distances, there's a reasonable chance the planet might become tidally locked, meaning its rotation is synched with its orbit so that only one face ever faces its host star. Under those circumstances, it's possible to form what at least one researcher termed an "eyeball Earth," with liquid water on the side facing the star, and ice frozen elsewhere.
Artist's conception of the "eyeball Earth," where a dim star melts the side of the planet that faces it.
Closer in, something different could occur. Greenhouse gasses or the equivalent of ocean circulation could carry heat from the star-facing side to the cooler, darker side, evening out the planet's temperature and creating a variety of possible habitable zones on the planet.
A survey to follow the survey
As Kepler continues to generate data, we'll almost certainly continue to see reports of bodies that are more Earth-like in size and orbit, along with estimates of whether they reside in the habitable zone. It's important to remember these estimates will tend to vary based on exactly which model of habitability the authors use. And they can never account for one-off factors—lack of water, highly reflective surface, or acid-belching volcanoes—that could either shift it out of the habitable zone or make it inhospitable to life even if it is.
All of which adds a big caveat to the numbers provided by Kepler, which suggest that we won't have to look too far to find an exoplanet the right distance from its host star. If we rush matters and build hardware that can only see a short distance, we'll only be getting a partial picture. To really get a sense about the prospects of life outside our Solar System, we need the equivalent of Kepler: a survey mission that can look at lots of planets, figuring out how common atmospheres are, how much light gets reflected back from them, and so on. That will require hardware that can resolve things far deeper than a few dozen light years.
A sense of how common these key features are would give us a far greater sense of the odds of finding a hospitable planet outside our solar system. And, if done right, it just may provide us with an indication of life in the process.
Here's hoping we'll see it happen while we're still alive.
i like it
ReplyDelete