QUOTE (elakdawalla @ Mar 10 2006, 07:50 AM)
![*](http://www.unmannedspaceflight.com/style_images/ip.boardpr/post_snapback.gif)
I'm a geologist, not an astronomer, so I only know what I have to know about spectroscopy...but...does it seem premature to anyone else to call things "the same color" when you only have views through two filters? That's a two-point spectrum. I understand that the filters on Hubble have probably been carefully chosen to make the most of differences at particular points on the electromagnetic spectrum, but still, two points isn't much of a spectrum, is it? Am I missing something?
--Emily
Well, with three we can tell the difference between roses, emeralds, and the blue sky of a summer day.
Not to trivialize the difference between two and three: there's a reason why tripods sell better than bipods.
I think the answer lies in looking at known spectra. Typically, you see either a flat line, a sloped line, or a "hump", with definite exceptions for absorption lines that send the spectrum potentially near to zero, before returning the the general trend. It's easy to avoid the likely absorption lines, since we know that these worlds might have CO2, CO, N2, CH4, and H2O ices on their surfaces and aren't likely to be beryllium-strontium-cobalt. Then the question is, how many samples do you need to distinguish a flat line from a slope or a hump? Two seems to do it if you can avoid the case where you sample two values on opposite sides of the hump.
Check out the visible spectrum of Pluto (left end of the top figure here)
http://ifp.uni-muenster.de/~sohl/images/pl...to_spectrum.jpgIt rises quite steeply through the visible. Two points could distinguish something like that from a flat line.
Of course, it's theoretically possible that something psychopathic happens to a spectrum between two samples, but when we can constrain the possible materials, we can virtually eliminate that possibility.