Titan, the largest moon of Saturn, imaged with the Simultaneous Differential Imager, on February 12th, 2004. The disk of Titan is 0.8 arcseconds across in this image, or about 2250 times smaller (in angular length) than the full moon. The colors in the image represent corresponding near-infrared filters, Red=1.575 um, Green =1.600 um, Blue = 1.625 um.
 |
 |
Above, left is one of the dozen or so raw images that went into making the three-color Titan picture at the top of the page, and shows how our SDI technique for finding planets works. Those white dots, by the way, are all bad pixels on our chip, and are removed during subsequent data-processing steps.
After light passes through the telescope optics (including the adaptive optics needed to make the image as sharp as possible), it enters a double Wollaston prism (see above right), which splits the beam into four identical beams (to within a few nanometers, anyway). Then, the set of four beams enters a quad-filter, giving us an image at 1.575 microns, one at 1.6 microns, and two at 1.625 microns. So an observation of an object gives us images in three colors simultaneously. In the case of this Titan exposure, the left two images of the moon are within the methane absorption feature, while the two on the right (upper at 1.6, lower at 1.575 microns) are outside of this feature.
Shown above is a spectrum of the first discovered brown dwarf, Gliesse 229B, with the approximate ranges of our three filters indicated. The vibro-rotational transitions of methand really cut out the light in these objects (brown dwarfs and young planets) redward of 1.6 microns. Since Titan has a decent amount of methane in its atmosphere, it appears noticably darker within the methane absorption band (left two images of the raw SDI frame). Visible light, by the way, has wavelengths between about 0.9 (red) and 0.4 (blue) microns.
And here's a theoretical spectrum of a 3 Jupiter mass planet at an age of 30 million years. This model was produced by David Sudarsky, and represents the current concensus on the characteristics of young substellar objects. The red, green, and blue curves show the exact filter transmission curves, and you can see how nicely they fit around the features of the planet. In particular you can see how much less flux you'll get from Filter 3 compared to Filter 2.