Astro Seminar: Finding brown dwarfs

19Feb07

To the astronomer, perhaps the single most helpful property of brown dwarfs is this: YOU CAN SEE THEM. Brown-dwarf-ologists are not obliged to resort to the observational craftiness of their planetologist friends, but can just, well, take a picture. At least sometimes.

To illustrate, here’s a fun exercise: find the brown dwarfs.

See? So very easy.

Okay, that was mean. The image above is an infrared view of the galactic center — each one of those tiny pixels of light is an entire star and might possibly be worth looking at. The fundamental question in searching for brown dwarfs is HOW DO YOU FIND ANYTHING IN SUCH A HUGE UNIVERSE? How do you separate the sheep from the goats on your brown dwarf day of judgement? Where do you even start?

You start by figuring out what, exactly, to look for. You want a set of brown dwarf characteristics that, in conjunction, will let you pick your dwarfs out of a sky of stars. Here’s Adam’s list:

1. Look in the near-infrared.
2. Look for “funny” colors.
3. Look for high proper motion.
4. Look in young star-forming regions.
5. Look around nearby stars

Let’s start with the first, shall we? Okay. In theory, an object that absorbs all radiation incident upon it — a “blackbody” — should re-emit that energy as a characteristic spectrum dependent only on the object’s temperature. Hotter objects emit more light at shorter wavelengths, and colder objects at longer wavelengths. While not perfect blackbodies, most stars are a reasonable approximation thereof. So really hot stars look bluish; medium-hot stars look yellow; cool stars look reddish; and cold little brown dwarfs look redder-than-red, a.k.a. infrared.

Brown dwarf surface temperatures run from around 3000K for young dwarfs to 700K or so for the oldest, coldest objects. 700 K is about 800 F. Your oven bakes bread at 400 F…700 K is pretty damn cold for a star.

Astronomers often plot a star’s magnitude against its temperature, as measured by its color. Here’s a diagram from Hipparcos, a survey program that took pictures of, um, MANY stars:

Brown dwarfs are very faint, and very red. See the tail of stars in the lower right, where the plot looks like a tornado? Follow that down off the chart for a few inches. Brown dwarfs are there.

The best way to get data on LOTS of stars at near-IR wavelengths is to use any of several recent all-sky surveys. There are three really useful ones:

1. DENIS, which took images in I, J, and K bands (centered at 0.8, 1.26, and 2.16 micron wavelengths, respectively). It covered the Southern sky, and has a lovely 355 million objects.
2. 2MASS, which took J, H, and K-band images (H is at 1.6 microns). 2MASS covered the entire sky and has something like 471 million sources
3. SDSS, which took u, g, r, i, and z-band images.

Once you have LOTS of stars, you can plot various sets of data. Color-color diagrams plot the difference between two sets of filters, so you might have J-K one one axis, and J-H on the other. Color-magnitude diagrams plot a difference on one axis and a straight-up single-filter magnitude on the other. The point of all this is that populations of similar stars will tend to cluster in a particular region on the plot, and stars with more unusual colors will make themselves known. Then you go AHA! and draw a big circle around them:

I love this population-analysis thing. It appeals to the same bit of brain that thinks stat mech is awesome.

Note that these are just some images I found on the 2MASS site and stuck some circles on. The location of brown dwarfiness is about right, I think, but I’m going by the sketches I have in my notes. I was considering asking Adam for his pictures, but… “Hi, can I blog your lecture slides?”…would that be weird? Maybe they’re classified, or something. If you were/are a professor, what would you think?

For a little more insight into this spectrum thing, consider this diagram. It’s showing (approximately) the normalized spectra of Vega, a bright star of spectral class A, and of a brown dwarf.

Vega emits much of its light in the optical region of the spectrum, which is why we see it as shining so brightly. The brown dwarf does a lot of emitting in the near-IR, but is fraught with absorption lines of this and that, so the spectrum ends up a little raggedy.

Now let’s superimpose a few filter bands on those two spectra. The V band is at the center of the optical range (V for visible?); J and K are infrared. Vega is very bright in V but much less so in J. The dwarf, on the other hand, is bright in J but dim in V. So on a plot of J-V, the dwarf would have a high value and Vega a low one.

Searching for brown dwarfs all over the sky has its advantages: you’ll end up with the dwarfs that are closest to the Sun, which means they’re more easily studied, and can be detected at fainter magnitudes and colder temperatures. Plus, you can use existing sky surveys. The disadvantages? Well, for one, brown dwarfs are RARE. For his dissertation, Adam looked at half of the entire sky, and found all of 10 T-dwarfs. Also, the sky surveys don’t let you constrain the parameters very well — you don’t know the distance of your object, or its mass, or its age…

One solution to the parameter problem is to look for dwarfs in star clusters. If the dwarf is, indeed, part of the cluster, you can place constraints on its age and distance. The biggest problem here is to make sure that what you’re looking at really is a brown dwarf. And for this, you look once again at the spectrum — specifically, you look for the signature absorption pattern of lithium. Stars with a mass higher than about 0.065 solar masses will quickly consume their lithium, and it will be entirely absent from their spectra. Brown dwarfs, on the other hand, won’t. So finding lithium absorption indicates that your object is below that mass limit, which makes it nice and small and probably brown (purple!) and therefore of interest. And there is much rejoicing.

Lecture ended at this point. I won’t be at seminar this week, but I’ll ask Anna or someone to give me a report.

Adam’s research pages have some nice brown-dwarf information, plus links to his papers.

This paper by Oppenheimer, Kulkarni, and Stauffer gives a really nice overview of some features of brown dwarfs, and explains just what makes a star a star, and a planet a planet, and why a brown dwarf isn’t either. I particularly like Figure 2, which shows the evolution of luminosity for objects on the borderlines, and explains each bump and wiggle.

P.S. The image of the galactic center is from 2MASS, about which I am supposed to say, “Atlas Image obtained as part of the Two Micron All Sky Survey (2MASS), a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation.”

The Hipparcos HR diagram is from http://orca.phys.uvic.ca/astrocourses/a120/A120/lab5.html.

The color-color and color-magnitude diagrams, minus the red circles, are from http://www.ipac.caltech.edu/2mass/releases/allsky/doc/sec1_2.html. They, too, are from 2MASS, which means you should go read that “joint project..funded by” blurb again.

Filed under: classes, seminar   |  5 Comments