In two talks in the fall of 2012, I used an incredible image produced by the orbiting Fermi Gamma-ray Observatory, reproduced here. My presentations were on black holes—particularly the supermassive black holes that lurk in the nuclei of nearly every large galaxy in the known Universe. (You can read summaries of these presentations both here and at Bowler Hat Science.) Many of these black holes are active, spewing out huge jets of light and matter that can stretch for millions of light-years. As galaxies go, the Milky Way is remarkably quiet, though, but the discovery in 2010 of two huge bubbles extending from the galaxy’s center gave rise to speculation that they emanated from past quasar-like activity from the black hole.

I used the Fermi image in my presentations for that reason, but to my discredit, I didn’t mention the alternative idea: that the bubbles were the product of star formation near the galactic center. Don’t get me wrong—I’m not blaming myself for hoping the black hole was the culprit. With gamma ray data alone, there was no easy way to distinguish between the two possible explanations for the bubbles, so it was reasonable for me to present the idea. In other words, I didn’t make it up.

However, follow-up observations using radio and microwave data are against the black hole hypothesis. The venerable Parkes radio telescope in Australia has been mapping the sky in polarized radio emissions; these observations turned up bubbles of gas of the same size and location as the Fermi data. Additionally, Wilkinson Microwave Anisotropy Probe (WMAP) data (aimed at studying the cosmic microwave background, but providing a map of the entire sky in microwave light to accomplish the task) showed strong evidence for the bubbles. However, the polarization—the orientation of the electric field of light—provided information about the magnetic fields within the bubbles. The higher resolution available to radio observation also turned up substructure: winding spirals of higher density. Those spirals, which make “ridges” in the contours of radio emission, are incredibly regular in distribution.

Together, these data are very suggestive of the following scenario. In the central 100 parsecs (330 light-years) or so of the Milky Way, the galaxy is producing new stars at a fairly rapid rate. That’s known from other observations, so it’s not controversial. Star formation pumps a lot of matter into its surroundings, in the form of stellar winds. These winds are largely plasma: the electrically neutral mixture of electrons and the nuclei from which they were stripped. However, electrons are a lot lighter, so they get accelerated to high velocities—in this case, 1000 kilometers/second, fast enough to escape the galaxy’s gravitational pull. (For reference, the speed of light is about 300,000 km/s.)

The star formation doesn’t occur everywhere equally: it’s concentrated in particular spots. As the galaxy rotates, therefore, the effect is to make the winds spiral…precisely the effect seen in the Milky Way’s bubbles. The magnetic fields from the intense star formation also permeate the bubbles, producing the unique polarization signal seen in the Parkes Telescope data. The rapid electron motion also kicks light from background sources up to higher energies (a process known as “inverse Compton scattering”), producing the gamma ray signature seen by the Fermi telescope.

In other words, the multiple types of observation led to a coherent, consistent explanation for the bubbles. The black hole scenario simply doesn’t hold up under the new data. I admit: I’m a little disappointed, especially since it makes my earlier talks incorrect in that detail. (Note that the black hole is still there! We have other strong evidence for its existence, including the motion of nearby stars.) However, that’s the way science goes sometimes: a favored explanation often falls to new evidence. All theories should be considered contingent and subject to revision in the face of new data. The process of discovery involves rethinking, and sometimes jettisoning one hypothesis in favor of another.

In this case, it doesn’t hurt that the correct explanation reveals something significant about the central portion of our galaxy—and possibly even why the black hole is so quiet. According to the study, gas flowing in toward the nucleus is being used to feed star formation, rather than making the black hole bright. In other words, the bubbles are a sign of the black holes quiescence. Isn’t that worth being wrong?

(I also wrote this paper up for Ars Technica. The post title is from They Might Be Giants.)