|Image Credit: NASA/JPL-Caltech/UCLA|
First, some quick background and definitions. Stars like the sun shine because they are giant nuclear fusion reactors, with most fusing hydrogen into helium. This process releases a lot of energy and can last for a long time - the sun's lifetime is about 10 billion years. In stars that are smaller (less massive) than the sun, the fusion reactions still occur, but at a lesser rate, because the star is cooler and the hydrogen a bit more lethargic. This trend continues until you get to stars that are about 8% of the sun's mass. Since these stars fuse hydrogen so slowly, they may actually survive for thousands of billions of years (i.e., trillions of years), even though they don't have as much hydrogen fuel as the sun does.
Objects less massive than 8% the sun's mass simply are too cold to perform hydrogen fusion. There are a few uncommon nuclear reactions that can occur and keep the object warm for a few million years, but those reactions quickly flicker and die. Without energy from nuclear reactions, the "failed star" will begin to cool off and slowly fade away. These objects are called brown dwarfs.
Smaller yet, at about 1.5% the mass of the sun, objects can never perform fusion reactions at all. Many astronomers refer to these tiny objects as planets, although there is no physical law that says these things must orbit stars (a condition some astronomers think is necessary to call something a planet). Jupiter, although a whopping big planet in our Solar System, is only 0.1% the mass of the sun, 15 times less massive than this limit. Although some books and people may refer to Jupiter as a "failed star", it never came close to doing any type of nuclear fusion and so being a brown dwarf, let alone sustaining nuclear fusion like stars do. Calling Jupiter a "failed star" is, in my opinion, like calling my house cat a tiger. It may look and act like a little tiger, but it's a different animal. Anyway, I digress.
Stars and brown dwarfs are classified based on their surface temperatures. When we take a spectrum of a star, we split its light up into its component colors. The different elements and molecules present at the surface of a star each have a sort of bar code in the spectrum, so we can tell what a star is made out of by looking at the spectrum.
But the elements we see depend not just on how much of that element is present, but also how hot the star is. For example, about 90% of the sun's atoms are hydrogen atoms, but the sun's spectrum is dominated by bar codes from elements like iron, magnesium and calcium, elements that make up less than 1% of the sun's atoms.
For stars hotter than the sun, like the bright star Sirius, we see mostly hydrogen in the spectrum. for stars even hotter yet, we see mostly helium. But all stars are nearly identical to the sun in composition -- it's the temperature that makes the huge differences in the spectrum.
About 100 years ago, a group of women astronomers at Harvard College Observatory devised a system for classifying the spectra of stars by using a letter for each type of spectrum, starting with "A", "B", and so on up to "Q". Each letter stood for a spectrum that was dominated by different elements. The system was developed before astronomers understood that the star's temperature was important. Once the temperature effects were figured out, the spectral types were simplified and re-arranged to the now-familiar "O B A F G K M" sequence. O stars are very hot, while M stars are the coolest true stars around.
Brown dwarfs were first discovered in the 1980s and 1990s. Young brown dwarfs are still hot from their formation and the tiny amounts of nuclear fusion that they can perform, and so have temperatures that give them M type spectra. But as astronomers discovered more and more brown dwarfs, they began to find some with cooler spectra that looked quite different from the spectra of M stars. In 1999, Caltech astronomer Davy Kirkpatrick organized these brown dwarfs into two new spectral types.
After careful consideration of possible names for these new spectral types, he and his collaborators settled on the letters "L" and "T", being two of the only three letters left that would not be confusing. L-type brown dwarfs, the next class cooler than M stars, show strong features of molecules like iron hydride and atoms such as sodium and potassium in their spectra.
As a brown dwarf cools further, methane will form in its atmosphere and become a strong feature in the spectrum. Since these spectra look much different, they are a different spectral type, the "T"-type brown dwarfs. Until recently, all known brown dwarfs were spectral type "L" and "T", which covered temperatures down to about 500 Kelvin (about 230 Celcius, or about 450 degrees Fahrenheit) -- much cooler than stars, but still fairly toasty by human standards. As a comparison, the planet Jupiter has a temperature of about -150 Celcius, or what my grandmother would call "boo cold".
The third remaining spectral type letter, "Y", Kirkpatrick suggested to be used for the coolest possible brown dwarfs, those with temperatures below about 500 K. At these temperatures, ammonia can form and become strong in the spectrum. While a few cool brown dwarfs that might have been "Y"-type had been discovered and described, there has been vigorous debate over whether these really were a different spectral class.
Then along came WISE. WISE is a telescope that looks in the infrared and is optimized to find things with temperatures close to room temperature (or colder). WISE has therefore found numerous asteroids and comets, and it was expected to find many brown dwarfs. Brown dwarfs don't give off much visible light, but shine in the infrared. Once the WISE team identified candidate cold brown dwarfs, they went to some of the largest existing ground-based telescopes like Magellan and Keck to get spectra. After all, if you are going to be claiming to find a new spectral type, you better have a spectrum to prove it.
The spectra tell the tale. There does indeed appear to be some ammonia visible in the spectra of these WISE brown dwarfs, and there are other substantial differences with the coolest T dwarfs. So, it seems very likely that these cool brown dwarfs are indeed "Y" dwarfs, fulfilling a prediction made over a decade ago. And the temperatures of these stars are around 300 to 500 Kelvin, or from room temperature up to a few hundred degrees. The coolest of the confirmed Y dwarfs, with the typical astronomically inscrutable name of WISE 1828+2650 (the green dot in the photo at the top of this post), has a surface temperature of about 80 degrees Fahrenheit, very comfortable for humans!
Don't start making vacation plans to visit WISE 1828+2650 yet, though. First, like all stars and brown dwarfs, WISE 1828+2650 is made entirely of gas and has no "surface" on which you could stand. And even if there were some sort of platform on which you could stand in the 80-degree atmosphere, the force of gravity would be 10 to 100 times that on Earth -- not exactly pleasant. And, as if that weren't bad enough, like all stars, WISE 1828+2650 is composed almost entirely of hydrogen and helium, so there wouldn't be enough oxygen to breathe.
Could there be cooler brown dwarfs out there? Quite possibly. We don't know exactly how fast brown dwarfs cool off, but small brown dwarfs can cool to 300 Kelvin in 5 to 10 billion years, less than the age of the Universe. "Planets" obviously can get much colder yet, like Jupiter at its frigid 125 Kelvin. But the oldest, largest brown dwarfs may not have had enough time to get this cold yet. So, the green dot in the picutre at the top of this post could be one of the coolest stars in the sky, in more ways than one.
Michael C. Cushing, J. Davy Kirkpatrick, Christopher R. Gelino, Roger L. Griffith, Michael F. Skrutskie, Amanda K. Mainzer, Kenneth A. Marsh, Charles A. Beichman, Adam J. Burgasser, Lisa A. Prato, Robert A. Simcoe, Mark S. Marley, D. Saumon, Richard S. Freedman, Peter R. Eisenhardt, & Edward L. Wright (2011). The Discovery of Y Dwarfs Using Data from the Wide-field Infrared Survey Explorer (WISE), accepted for publication in The Astrophysical Journal, arXiv: 1108.4678v1