What is a spectrum?

Back in the Enlightenment era, a confused young man named Isaac Newton wondered whether the colors we see are an intrinsic property of light or whether they occur in the human eye.  To test this, he did various things, including wedging a bodkin between his eyeball and bone.

Although the bodkin was not terribly revealing, a prism was.  A simple piece of wedge-shaped glass, when placed in the path of a beam of white light, divided the light into a rainbow of colors.  Because these colors were clearly a part of the white light, but could only be revealed with a device like a prism, Newton called the rainbow a spectrum, the Latin word for “apparition.”

How is a spectrum produced?

While modern-day astronomers sometimes use a prism, tools for spreading light into a spectrum have become more advanced.  A diffraction grating (interactive demo) is a regularly etched surface that can separate a beam of light into its component colors, or wavelengths.  The tightly spaced etchings create interference patterns in the incoming waves of light, with the result that different wavelengths of light reflect constructively at different angles.  Thus, the different wavelengths of light travel to different regions of space, creating a spectrum.

In most of the work my team does, our light-dividing tool is an echelle spectrograph.  “Echelle” is the Italian word for stairs, and this word makes sense because our tool is a diffraction grating that looks like a staircase.  The angle of the staircase determines how much the spectrum will spread: the steeper the staircase, the more two wavelengths get split apart.  An image from one of my favorite echelle spectrographs is shown below.


This is a spectrum taken with the Levy spectrograph on the Automated Planet Finder telescope at Lick Observatory, CA. The spectrum is produced by combining an echelle grating (which separates light in the left-right direction in this image) with a cross-disperser (which separates light in the up-down direction in this image). The pixels have been tinted to illustrate which parts of the image correspond to which wavelengths of light, but in reality, the image is grayscale (we simply count the number of photons in each pixel and rely on calibration methods to figure out the wavelength of light falling on each pixel).  The sharp, dark, horizontal lines are produced by the optical setup and are meaningless, but the soft, dark, vertical lines correspond to atoms and molecules interacting with the starlight.  Many of these atomic features are in the atmosphere of the star.  They reveal information about the temperature, gravity, and composition, and motion of the outer layers of the star.  However, the repetitive patterns of sharp, dark lines near the top of the image (red light) are caused by oxygen molecules in Earth’s atmosphere.  The large-scale, faint dark squiggles at the top of the image are interference patterns from imperfections in the optical surfaces.

What can we learn from a spectrum?

Dividing a source of light into its component wavelengths can reveal many things about the light source.  When Joseph von Fraunhofer did this for the sun by placing a prism behind a telescope, he discovered that the white light of the sun does not split into a perfectly even rainbow.  Some wavelengths of light seemed to be missing from the sun’s spectrum.  These lines are now called the Fraunhofer lines.  They correspond to wavelengths of light at which particular elements in the upper atmosphere of the sun absorb sunlight.  A few notable elements in the Fraunhofer lines are sodium (the same color in yellow street lamps is diminished in sunlight), magnesium, iron–and of course, good old hydrogen and helium, which we now know are the most common elements in the sun thanks to the pioneering work of the twentieth century astronomer Cecilia Payne-Gaposchkin.

In addition to probing the chemical constituents of stars from afar, the spectrum can reveal other fundamental properties.  Annie Jump Cannon found that the strength of the hydrogen absorption lines could be used to classify stars, and her classification system was later linked with the temperatures of stars.

Nowadays, we can interpret the temperatures, atmospheric pressures, and gravities of stars, their surface rotation speeds, their elemental, ionic, and molecular compositions, their speeds toward or away from us, and the speed of outflowing stellar winds from the spectra of stars.  With the aid of polarimetry, spectra can also probe the magnetic fields of stars.  (Did I miss your favorite physical property?  Let me know in the comments.)  The wealth of stellar physics that we can mine from spectra is mind-blowing.

Spectra can also be obtained and studied for objects that are not stars.  The spectrum of a galaxy is a combination of the spectra from its light sources, which are primarily stars.  Spectra have also been taken of the hot disks of gas that slowly spiral into black holes, of the cool gas that eventually coalesces to form stars, and of the circumstellar disks around young stars where planets will soon form.  In studying exoplanets, we can take the spectrum of starlight that has passed through or reflected off a planet atmosphere, revealing the planet’s atmospheric chemistry and temperature.

As my Introduction to Astronomy students often marvel, it’s amazing what you can learn from a scrawny drop of starlight.  The trick is in handling the light in a way that extracts useful information.