Stellar classification

• 1. Introduction

  ›

  The Sun carries a short label that astronomers write as G2V. Those three characters say it is a main-sequence star with a surface temperature of around 5,800 K. Nearly every star whose light can be studied receives a code like it. The rules that hand out those codes are called stellar classification, and they work by taking starlight apart. Yet the code hides a puzzle. The letters for stars, from hottest to coolest, run O, B, A, F, G, K, M, an order that looks scrambled. Why did astronomers settle on such an arrangement? How can a thin stream of light reveal a star's temperature, its size and its chemistry? And what happened when the people who wrote the labels discovered, years later, what the labels were really measuring?

  38395013

• 2. Reading the Rainbow

  ›

  Hold a prism up to starlight and the beam fans out into a rainbow. A diffraction grating does the same job. Crossing that rainbow are spectral lines, and each line points to a particular chemical element or molecule. A stronger line means more of that substance is present.

  This sounds like a simple ingredient list, but mostly it is not. The strength of the lines shifts chiefly with the temperature of the star's photosphere, its glowing surface layer. Only in some cases do the lines reflect true differences in abundance.

  That is why a spectral class works as a compact code. It mainly summarizes the ionization state of the atmosphere, which gives an objective measure of the photosphere's temperature.

  Lines are not the only route. Other modern systems, such as the UBV system, rely on color indices, which are measured differences between color magnitudes. Labels like U-V or B-V name the colors passed by two standard filters, for example ultraviolet, blue and visual.

  The sorting of spectra into classes began in the 1860s with Angelo Secchi, and his first scheme had only a handful of groups.

  48359382

• 3. The Harvard Computers

  ›

  By 1866, Angelo Secchi had divided stellar spectra into three classes. His first class held white and blue stars with broad hydrogen lines, such as Vega and Altair. The second held yellow stars with weaker hydrogen and clear metal lines, including Arcturus and Capella. The third covered orange to red stars with complex bands, such as Betelgeuse and Antares. In 1868 he found carbon stars and gave them a fourth class. In 1877 he added a fifth for stars with emission lines.

  The next leap came at Harvard. After the death of her husband, Mary Anna Draper began funding the Harvard Plate Stacks and the study of those plates at the Harvard College Observatory. The observatory's director, Edward C. Pickering, hired pioneering women astronomers who became known collectively as the Harvard Computers. In 1890 came the first edition of The Henry Draper Memorial Catalogue of Stellar Spectra.

  Williamina Fleming classified most of the spectra in that first edition. She is credited with classifying over 10,000 stars and discovering 10 novae and more than 200 variable stars. Working with Pickering, she separated 17 classes by the strength of hydrogen lines. Class A showed the strongest hydrogen absorption, while class O showed virtually none, so the alphabet tracked a steady fade in hydrogen.

  Antonia Maury took a different path in 1897. She placed the narrow-lined Orion subtype of Secchi's first class ahead of the rest, which put the modern type B before type A. She numbered her types with Roman numerals from I to XXII instead of using letters. Her catalogue, Spectra of Bright Stars Photographed with the 11 inch Draper Telescope as Part of the Henry Draper Memorial, held 4,800 photographs and her analyses of 681 bright northern stars. It was the first time a woman was credited in an observatory publication.

  In 1901, Annie Jump Cannon went back to letters. She kept only O, B, A, F, G, K, M and N, with P for planetary nebulae and Q for peculiar spectra. For stars between two types she wrote labels such as B5A, halfway from B to A. By 1912 those labels had become B5, F2 and their kin, which is essentially the modern Harvard form. The order of her letters stayed fixed, but the explanation for that order arrived only in the 1920s.

  4647485761

• 4. Labels That Mislead

  ›

  In the 1920s, the Indian physicist Meghnad Saha built a theory of ionization by stretching ideas about molecules splitting apart to cover atoms losing electrons. He applied it first to the solar chromosphere and then to stellar spectra. Harvard astronomer Cecilia Payne then showed that the O-B-A-F-G-K-M sequence is a sequence in temperature. Suspicion of this had existed since about 1914, when the first Hertzsprung-Russell diagram took shape.

  Because the letters were chosen before anyone understood that, placing a spectrum into a subtype such as B3 or A7 rests on largely subjective estimates of line strength. The subtypes therefore do not divide into even steps that a formula could describe.

  The vocabulary carries a second fossil. Astronomers call hot stars early and cool stars late. In the absolute sense, early means O or B and perhaps A, while an unqualified late means K and M. In the relative sense, early K might be K0 to K3, and late G might be G7 to G9. The words descend from a late nineteenth century model in which stars shone by gravitational contraction through the Kelvin-Helmholtz mechanism and cooled as they aged. That model gave the Sun an age far shorter than the geological record suggests. The discovery that stars run on nuclear fusion ended the model, but the words stayed.

  Color descriptions mislead as well. The conventional ones consider only the peak of a star's spectrum, yet stars radiate across all of it, and all spectral colors combined look white. So in typical viewing there are no green, cyan, indigo or violet stars. Yellow dwarfs such as the Sun look white, and red dwarfs look a deep yellow or orange. A brown dwarf would not look brown at all; to a nearby observer it would appear dim red or grey-black.

  A refinement in 1943 would add a second dimension to the labels, built from the width of the lines.

  1920184410

• 5. Adding a Second Dimension

  ›

  In 1943, William Wilson Morgan, Philip C. Keenan and Edith Kellman of Yerkes Observatory introduced a system that went beyond temperature. It is called the Yerkes, MK or MKK system, after the authors' initials. In 1953, after revisions to the standard stars and the criteria, it was renamed the Morgan-Keenan classification, and it remains in use today. Harvard's scheme had judged surface temperature alone, while the new one also used lines sensitive to surface gravity, which relates to luminosity.

  The physics comes down to pressure. Denser stars with higher surface gravity show more pressure broadening of their spectral lines. A giant has a far larger radius than a dwarf of similar mass, so the gravity and pressure at its surface are much lower. Its lines come out narrower. That lets an astronomer assign a luminosity class from the spectrum alone.

  The classes run from the top. Class 0 or Ia+ holds hypergiants, and class I holds supergiants, further split into Ia, Iab and Ib. Class II is bright giants, class III normal giants, and class IV subgiants. Class V is the main sequence, and sd or VI marks subdwarfs. Class D or VII is for white dwarfs, though that last numeral is now rarely used. Arcturus is K0III, a normal giant, while Achernar is B6Vpe, a dwarf on the main sequence.

  Marginal cases get special symbols. A slash means a star is one class or the other, and a hyphen means it sits between two. A star labeled A3-4III/IV lies between A3 and A4 and is either a giant or a subgiant. Fractions are allowed too, as with Mu Normae, which is O9.7.

  Lower-case letters after the type flag oddities. The star 59 Cygni is B1.5Vnne, meaning a B1.5V spectrum with very broad absorption lines and certain emission lines.

  The whole scheme is taxonomic, like sorting species in biology. Each category is defined by one or more standard stars, and those standards stretch from the rarest hot giants to the most common dim dwarfs.

  2324222125

• 6. The Seven Letters

  ›

  The traditional mnemonic for the order is "Oh, Be A Fine Guy/Girl: Kiss Me!" Astronomy courses and organizations have held contests for rivals, but the original is still the most popular.

  Class O is the rarest of all. About 1 in 3,000,000 main-sequence stars in the solar neighborhood belongs to it. These stars are extremely luminous and radiate mostly in the ultraviolet. Their stellar winds can reach 2,000 km/s, so they do not hold on to extensive atmospheres. They burn through hydrogen so fast that they are the first stars to leave the main sequence. When the MKK scheme appeared in 1943, it used only O5 to O9.5. It was extended to O9.7 in 1971 and to O4 in 1978.

  Regulus is the closest B-type star, at around 80 light years. B stars live only briefly, so they seldom wander far from the OB associations where they formed. The Orion OB1 association fills a large part of a spiral arm of the Milky Way and includes many of the brighter stars of Orion. The Be stars, which show hydrogen lines in emission, are thought to rotate at a curiously rapid rate.

  Vega, Sirius A, Deneb and Fomalhaut serve as standards for class A. These white or bluish-white stars have hydrogen lines that peak at A0, and they are among the more common naked-eye stars. Nine of them lie within 15 parsecs. Procyon A is the F-type star within 20 light years, a class whose lines of ionized calcium grow stronger.

  Chara and Muphrid are standards for class G, with 21 stars of that class within 10 parsecs. The class also contains the Yellow Evolutionary Void. Supergiants often swing between blue O or B and red K or M, and they do not linger in the unstable yellow stage between.

  Orange class K accounts for about 12 percent of main-sequence stars nearby. It includes the giant Arcturus, the orange dwarf Alpha Centauri B and the hypergiant RW Cephei.

  Class M is by far the most common. Its main-sequence members, the red dwarfs, are so dim that none is bright enough to see unaided under ordinary conditions. The brightest, Lacaille 8760, has a magnitude of 6.7, just past the usual limit of 6.5. Yet the class also holds some of the largest supergiants in the Milky Way, including VY Canis Majoris, VV Cephei, Antares and Betelgeuse.

  Objects that fit nowhere on this ladder forced astronomers to invent more letters.

  1415828476

• 7. Beyond the Letters

  ›

  The Wolf-Rayet stars, class W or WR, were once counted among the O stars. Their spectra lack hydrogen lines and show broad emission lines of highly ionized helium, nitrogen, carbon and sometimes oxygen. Most are thought to be dying supergiants whose hydrogen layers were blown away by stellar winds, exposing hot helium shells.

  At the cool end, the letters L, T and Y were created for infrared spectra. Class L got its letter because L was the remaining letter alphabetically closest to M. These dwarfs are very dark red and brightest in the infrared. Class T dwarfs have surface temperatures between about 550 and 1300 K, and methane is prominent in their spectra. Brown dwarfs cool as they age, passing from M through L and T to Y.

  As of August 2013, a total of 17 objects had been placed in class Y. The coolest known, WISE 0855-0714, is about 250 K and has a mass of just seven times that of Jupiter. Y dwarfs span 9 to 25 Jupiter masses, so they straddle the 13 Jupiter-mass deuterium-fusion limit that marks the current division between brown dwarfs and planets.

  Carbon stars have their own class, C, which absorbed the older R and N classes. They are mostly red giants near the end of their lives, with excess carbon in their atmospheres. Class S stars form a continuum between M stars and carbon stars, with strong zirconium oxide bands. The designation S10 has been used for only one star, Chi Cygni, at an extreme minimum.

  White dwarfs get class D, for degenerate. Its letters, such as DA for hydrogen-rich and DB for helium-rich, describe the composition of the atmosphere and are unrelated to the other letters. A number follows, a rounded form of 50400 divided by the effective temperature, as in DA1.5 for IK Pegasi B.

  Some remnants cannot be placed at all. A black hole emits no visible light of its own, so it would not appear on the Hertzsprung-Russell diagram. For neutron stars, a scheme with types I, II and III has been proposed. Neutrinos carry off so much heat that an isolated neutron star falls from billions of kelvin to around a million within a few years.

  The same sorting now guides a practical question: which stars could plausibly shelter life.

  9697343585

• 8. Searching for Life

  ›

  Humans know of exactly one star that hosts life, the Sun, which is rich in heavy elements and varies little in brightness. It is also unusual because it has no stellar companion, unlike many systems. With a sample of one, astronomers lean on Earth as a guide.

  Stars more than 1.5 times the Sun's mass, which means types O, B and A, age too quickly for advanced life to develop. At the other extreme, dwarfs of less than half the Sun's mass, type M, are likely to tidally lock planets in their habitable zones. Many astronomers still model red dwarf systems because of their sheer numbers and longevity.

  That leaves a middle band. NASA's Kepler Mission searches nearby main-sequence stars that are less massive than type A and more massive than type M. This makes F, G and K dwarfs the most probable hosts. K stars draw special interest because they pair a broad habitable zone with much lower harmful emission and long lifetimes.

  9291943813