4 ms·

It's mind-bendingly cool that people can figure out the composition of an atmosphere without actually being close to that atmosphere. Had not heard of dimethyl

It's mind-bendingly cool that people can figure out the composition of an atmosphere without actually being close to that atmosphere. Had not heard of dimethyl sulfide before. That's a good keyword to know.

julienchastang 3y ago
From a single pixel of light! Indeed mind blowing.
- treyd 3y ago
  That's somewhat misleading though. It's not really a pixel in that there's much more information available than, basically, 3 integers between 0 and 255. There's 4 instruments on the telescope that collect 4 different chunks of IR spectra, and there's very precise and granular intensity values for light received around any given wavelength. Much more detail than a "pixel" has in the typical sense we use them. It's not a lot of information, not nearly enough to identify surface features on an exoplanet, but it's very useful data if you're trying to identify likely chemical composition of bodies or how hot clouds of gas are.
  - saltcured 3y ago
    It's a crude analogy, but I like to tell people that a spectroscope is more like a highly directional microphone to listen to molecules than it is like any normal camera...
  - kjkjadksj 3y ago
    If you were at such distance from the earth, what would be resolvable?
finite_depth 3y ago
This is an easier problem than it probably seems. Atmospheres are relatively low-density gas, which means they produce an absorption spectrum - a chemical fingerprint that identifies most atoms and molecules quite reliably. It's such a good fingerprint that it was used to discover several chemical elements, most notably helium (observed in the Sun's spectrum before it was known on Earth).
dghughes 3y ago
For emission spectroscopy has been a thing since 1859 when Gustav Kirchhoff figured it out but for stellar emission spectroscopy it was Joseph von Fraunhofer in early 1800s? Although 1866, Pietro Angelo Secchi may have also discovered stellar emission spectroscopy 1848. An amazingly long time ago.

julienchastang 3y ago

From a single pixel of light! Indeed mind blowing.

treyd 3y ago
That's somewhat misleading though. It's not really a pixel in that there's much more information available than, basically, 3 integers between 0 and 255. There's 4 instruments on the telescope that collect 4 different chunks of IR spectra, and there's very precise and granular intensity values for light received around any given wavelength. Much more detail than a "pixel" has in the typical sense we use them. It's not a lot of information, not nearly enough to identify surface features on an exoplanet, but it's very useful data if you're trying to identify likely chemical composition of bodies or how hot clouds of gas are.
- saltcured 3y ago
  It's a crude analogy, but I like to tell people that a spectroscope is more like a highly directional microphone to listen to molecules than it is like any normal camera...
- kjkjadksj 3y ago
  If you were at such distance from the earth, what would be resolvable?

finite_depth 3y ago

This is an easier problem than it probably seems. Atmospheres are relatively low-density gas, which means they produce an absorption spectrum - a chemical fingerprint that identifies most atoms and molecules quite reliably. It's such a good fingerprint that it was used to discover several chemical elements, most notably helium (observed in the Sun's spectrum before it was known on Earth).

dghughes 3y ago

For emission spectroscopy has been a thing since 1859 when Gustav Kirchhoff figured it out but for stellar emission spectroscopy it was Joseph von Fraunhofer in early 1800s? Although 1866, Pietro Angelo Secchi may have also discovered stellar emission spectroscopy 1848. An amazingly long time ago.