Spectral Classification of Stars

When the light from a star is divided into its component colors using a spectrograph, it appears as a continuous band of colors, broken up by dark, narrow lines. These lines are created by atoms and ions (atoms missing one or more electrons) in the outer layers of a star's atmosphere. These layers absorb light at specific wavelengths, which are unique for each type of atom or ion. Atomic physics predicts the positions and intensities of these lines, called absorption lines, based on the temperature and composition of the star. Thus, the number, strengths, and positions of these lines vary from star to star.

The first stellar spectra were observed in 1814, long before the atomic physics that creates them was understood. In an attempt to understand the processes which formed the spectra, similar stars with similar spectra were grouped together in the hopes that stars which were alike would produce similar spectra. In 1863, Father Angelo Secchi made one of the first attempts at trying to classify stars, when he divided stars into two groups based on their spectral lines. He eventually extended this categorization, dividing more than 4,000 stars into four classes.

The basis of our current system of classification of spectral types began in the late 1800s at the Harvard College Observatory, under the direction of Professor Edward The light intensity curves of a stars indicates their temperature. The y axis depicts intensity, the x axis depicts increasing photon wavelengths. White dwarfs are hot; red giants are cooler stars. Illustration by Argosy. The Gale Group. C. Pickering. Williamina P. Fleming initially classified 10,000 stars using the letters of the alphabet to denote the strength of their hydrogen absorption lines, with A being the strongest, followed by B, C, etc. At the time, she did not know that these lines were due to hydrogen, but since they were visible in almost all stellar spectra, they provided a convenient means by which to organize her data.

Several years later, the classifications were reordered to be in what we now know to be the order of decreasing temperature: O, B, A, F, G, K, M, in order to have a smooth transition between the class boundaries. This reordering was done primarily by Annie Jump Cannon, also at the Harvard Observatory, in preparing the Henry Draper catalog of 225,000 stars. She also further subdivided each class into as many as ten subclasses, by adding the numbers 0 through 9 after the letter, to account for changes within a class. This spectral classification scheme was formally adopted by the International Astronomical Union in 1922, and is still used today.

It was not until 1925 that the theoretical basis behind the ordering was discovered. At first, scientists believed that the strength of the lines directly determined the amount of each element found in the star, but the situation proved more complex than that. Most stars have very similar compositions, so the strength of hydrogen (and other) lines in the spectrum is not a measure of the makeup of the star. Instead, it is a measure of temperature, as a result of the atomic physics processes occurring in the star. At relatively low temperatures, the gas in a stellar atmosphere contains many atoms, (and even some molecules), and these produce the strongest absorption lines. At higher temperatures, molecules are destroyed and atoms begin losing electrons, and absorption lines of ions begin to appear. More and more ionization occurs as the temperature increases, further altering the pattern of absorption lines. Thus, the smooth sequence of line patterns is actually a temperature sequence.

Another of the early classification workers at Harvard, Antonia Maury, noted that certain dark lines (absorption lines) in stellar spectra varied in width. She attempted a classification based partially on line widths, but this was not adopted by Annie Cannon in her classification, and was not used in the Henry Draper catalog. But Maury's work laid the foundation for the subsequent discovery that the line widths were related to stellar size: very large stars, now called giants or supergiants, have thin lines due to their low atmospheric pressure. These stars are very luminous because they have large surface areas, and so line width was eventually recognized as an indicator of stellar luminosity.

In 1938, W. W. Morgan at the Yerkes Observatory added a second dimension to the classification scheme, by using the luminosity of the star as an additional classifying feature. He used roman numerals to represent the various types of stars. In 1943, the MKK Atlas of Stellar Spectra (after Morgan, P. C. Keenan, and E. Kellman) was published, formalizing this system. Approximately 90% of stars can be classified using the MKK system.