Looking Stellar: How We Classify Stars

Last week’s post about the stars of the Summer Triangle had me dipping my toe into the seemingly odd system that is modern stellar classification, and it made me realize that it’s worthy of its own post. How exactly do we classify stars? And why the absolute heck do we use what appears to be a totally random set of letters to do it?

This one includes history, science, and women scientists being completely awesome, so it’s right up my alley. Join me as we examine just how we categorize those lovely lights in the sky.

Where Those Letters Come From

Image

As a species we didn’t get try to get precise about classifying stars until we were able to start looking at their spectra (as a quick reminder, a star’s spectrum is what you get when you pass light from the star through a spectroscope, which will break it up into all its components the same way a prism splits sunlight into a rainbow. The various elements and molecules making up a star will leave distinctive patterns of lines in its spectrum). 

The first person to attempt to create a systematic way of classifying stars based on their spectra was a guy named Angelo Secchi in the 1860s. He focused primarily (though not exclusively) on stars’ hydrogen lines to divvy them up into five classes marked by Roman numerals.

Of course Secchi was just eyeballing the stars’ spectra one at a time. In 1872 an American astronomer named Henry Draper managed to photograph a stellar spectrum for the first time. Recording the spectra, of course, allows them to be directly compared to each other.

Image

Draper died young in 1882, but his widow gave a lot of money to Harvard Observatory, then among the most preeminent astronomical facilities in North America, to create a whole catalog of images of stellar spectra. Then it fell to the Observatory’s famous group of female astronomers, known as Pickering’s Computers (after the Observatory Director who hired them), to classify them.

While this group included several women who made groundbreaking contributions to astronomy, it was one in particular, Williamina Fleming, who led the charge in using these spectra to create a refined version of the Secchi system. It was still primarily based on hydrogen lines and was published in 1890. 

 Known as the Draper System, each of the five Secchi classes were subdivided into smaller groups, each of which was assigned a letter. At this particular point the lettering system still made sense: Secchi class I was subdivided into classes A, B, C, and D in the Draper Classification System, for instance. The letters went up to Q.

Who Needs Alphabetical Order?

Then in the early 20^th century came the great Annie Jump Cannon. Another one of Pickering’s Computers, she took the Draper System and just…completely tore it apart and created something new out of it. Cannon focused her system on spectral lines that are affected by surface temperatures, rather than hyper-focusing on a particular element.

Image

She found that, looked at in this light (ha, spectrum joke), many of the existing letter classifications could be mushed together, leaving the letters A, B, F, G, K, M, and O. And then, if you’re going to focus on temperature going from hottest to coolest, those letters need to be reorganized as O, B, A, F, G, K, M. The mnemonic I was taught for this was “Oh, Be a Fine Girl/Guy, Kiss Me”, which is, let’s face it, a touch on the creepy side by modern standards. Each of these classes is then divided into ten subclasses marked 0-9, with 0 being the hotter end and 9 being the cooler.

Criminally this system is not known as the Cannon System, but the Harvard System. It forms the base of the system we use today, known as the Morgan-Keenan (MK) System, sometimes called the Yerkes System. Essentially this is just Cannon’s method with one extra element, a Roman numeral (yes, we’re back to those), sometimes accompanied by lowercase letters, that serves as a way of stating what stage of the star’s life it’s in (i.e. supergiant, regular main sequence, etc.).

Stars of the Show

Let’s look at those non-alphabetical star types! Note that for this section we’re looking specifically at main sequence stars, ones in the middle of their lives. We’ll start with the O stars. These are the ridiculously giant, honkin’ bright, super blue stars that burn insanely fast, die young, and blow themselves (and anything around them) to smithereens when they go.

Their surfaces are 60,000 F (about 32,000 C) at minimum and you could pack over 250 Suns inside the smallest of them. There are, frankly, not a lot of these around. Making a star this hot is hard, and then they just don’t stick around for very long. The biggest only make it a couple of million years at most before blowing up. Two of Orion’s three belt stars, Alnitak and Mintaka, are O-types.

B-type stars are also not exactly low key. The minimum surface temp for a star to qualify as a B is still over 17,500 F (about 10,000 C). The smallest of them could fit 25 Suns inside, which you’ll notice is a lot less than the 250 we had with the O-types, but we’re still talking about 25 Suns. 

Image

These ones can live for around 100 million years. Of course, that’s still not long as astronomers measure things (birds have been around for longer than 100 million years, just to give you a point of comparison). Some B stars are massive enough to supernova when they die, but not all. Probably the most famous B-type star in our sky is Rigel, Orion’s bright right knee star, though the remaining belt star, Alnilam, is also a B.

The A class is where the temperature and size ranges for the star classes begin to narrow, with A-types having surface temps anywhere from 12,500-17,500 F (roughly 7,000-10,000 C). The smallest A-type star can fit around 11.5 Suns inside and stars of this class can expect to pass the billion-year mark in their lifetimes. The brightest stars in Earth’s night sky, Sirius and Canopus, are A-types (as well as, if you remember from last week, Vega, Deneb, and Altair).

F-types range in temp only from about 10,300-12,500 F (about 5,700-7,000 C). The smallest F-types can fit less than ten Suns inside them and have lifespans around 5 billion years. The brightest F-type in our sky is Procyon, not far from Orion in the constellation Canis Minor, but the most famous has to be the North Star, Polaris.

You probably have noticed we’ve been getting more and more Sun-like. Well here we are at the G-type stars, which include our beloved Sun! These stars can be slightly bigger or slightly smaller than the Sun and can have a surface temp as low as 9,000 F (about 5,000 C). Fortunately for us these stars tend to stick around for about 10 billion years. 

Image

If you want the full MK System classification for our Sun, it’s a G2V star, indicating a hot G star in mid lifespan. It’s also one of the few easily visible G-type stars in our sky. Rigel Centaurus in the Southern Hemisphere and Capella in the Northern are really the only other ones of note.

Moving on to the cooler, smaller classes, K-types can be as cool as 6,500 F (about 3,600 C). You could fit about 1.5 of the smallest K stars inside of our Sun. We suspect they can live to be about 50 billion years old or so, but since this is well older than the age of the universe the earliest K-type stars still haven’t died.

This is a place where the distinction that I’m talking about main sequence stars is important, because the brightest K-type star in Earth’s sky is Arcturus. Arcturus is gigantic, waaaay bigger than the Sun, because it’s dying and therefore is no longer on the main sequence. It has swelled up into a giant, which has also caused its outer layers to cool into the K-type temp range. Before leaving the main sequence to start dying, Arcturus was probably a G or F-type.

Then there’s the wee little Ms. These are downright cold for stars, as cool as 3,600 F (about 2,000 C) and can get so small that it makes more sense to compare them in size to Jupiter than to the Sun. They can live, potentially, for hundreds of billions of years. 

Image

Again, that’s main sequence stars because the most famous M-types in our sky are the dying supergiants Betelgeuse and Antares. Main sequence M-types are so faint that we can’t see any without the use of telescopes, but our nearest stellar neighbor, Proxima Centauri, is an M5.5Ve

These are not actually the only stellar classifications. As we learn more about space, we’ve added a few additional letters. There is, for instance, an L class which covers the murky range of things that might be stars or might be brown dwarfs. T-types and Y-types could be brown dwarfs or gigantic free-floating planets. There are a few other letters that indicate very specific types of stars at very specific stages of their lives as the system drills down into more and more detail.

The Learn’d Astronomer

And being able to easily outline those details is key because accurately classifying things is very important for scientific endeavors! For an astronomer, knowing that Sirius is an A1V star instantly tells a lot about it—its size, its brightness, where it is in its lifetime, all in a single small designation.

But that’s the terrific thing about astronomy, it contains both deep scientific truths and unmatched natural beauty. I personally love the nitty gritty of systemic minutiae and knowing things like Betelgeuse is an M2Iab star. But it also makes me think of the poem “When I heard the learn’d astronomer” by Walt Whitman. Sometimes I enjoy reveling in the fact that Arcturus’s advanced age makes it a K1.5III-type star. And sometimes “I wander’d off by myself, in the mystical moist night-air, and from time to time, look’d up in perfect silence at the stars.”