Q&A: Understanding the Fermi Bubbles

In this Kavli Foundation Q&A, three astrophysicists who discovered two enormous and unexpected structures radiating from the center of our galaxy discuss what these mysterious bubbles can tell us about the history of the Milky Way and how they could help in the search for dark matter.

Compared to other galaxies, the Milky Way is a peaceful place. But it hasn't always been so sleepy.

galactic haze and Fermi bubbles

A pair of galactic bubbles can be seen in observations from NASA's Fermi Gamma-ray Telescope (blue) and ESA's Planck mission (red and yellow).

But it hasn't always been so sleepy. In 2010, a team of scientists working at the Harvard–Smithsonian Center for Astrophysics discovered a pair of “Fermi bubbles” extending tens of thousands of light-years above and below the Milky Way’s disk. These structures are enormous balloons of energetic gamma rays emanating from the center of our galaxy. They hint at a powerful event that took place millions of years ago, likely when the black hole at the center of our galaxy feasted on an enormous amount of gas and dust - perhaps several hundreds or even thousands of times the mass of the sun. But exactly how the bubbles formed, and the exact story they can tell us about the history of our galaxy, remains a mystery.

In Sky & Telescope's April 2014 issue, you can read about the decade-long odyssey launched by this discovery.

Now, courtesy of The Kavli Foundation, Sky & Telescope is featuring an in-depth Q&A with three astrophysicists who discovered the Milky Way bubbles, fresh from their Rossi Prize lecture at the winter meeting of the American Astronomical Society. They spoke with The Kavli Foundation about ongoing attempts to understand the cause and implications of these unexpected and strange structures, as well as ways in which they might help in the hunt for dark matter.

The following is an edited transcript of their roundtable discussion. The participants have been provided the opportunity to amend or edit their remarks.

THE KAVLI FOUNDATION: When the three of you discovered Fermi bubbles in 2010, they were a complete surprise. No one anticipated the existence of such structures. What were your first thoughts when you saw these huge bubbles – which span more than half of the visible sky – emerge from the data?

DOUGLAS FINKBEINER: How about crushing disappointment? There seems to be a popular misconception that scientists know what they’re looking for and when they find it, they know it. In reality, that’s often not how it works. In this case, we were on a quest to find dark matter, and we found something completely different. So at first I was puzzled, baffled, disappointed and confused.

We had been looking for evidence of dark matter in the inner galaxy, which would have shown up as gamma rays. And we did find an excess of gamma rays, so for a little while we thought this might be a dark matter signal. But as we did a better analysis and added more data, we started to see the edges of this structure. It looked like a big figure 8 with a balloon above and below the plane of the galaxy. Dark matter probably wouldn’t do that. At the time, I made the tongue-in-cheek comment that we had double bubble trouble. Instead of a nice spherical halo like we would see with dark matter, we were finding these two bubbles.

TRACY SLATYER: I called a talk on the Fermi bubbles “Double Bubble Trouble” – it has such a nice ring to it.

FINKBEINER: It does. After my first thought - “Oh darn, it’s not dark matter” - my second thought was, “Oh, it’s still something very interesting, so now let’s go find out what it is.”

SLATYER: At the time, Doug, you told me something along the lines of “Scientific discoveries are more often heralded by ‘Huh, that looks funny’ than by ‘Eureka!’” When we first started seeing the edge of these bubbles emerge, I remember looking at the maps with Doug, who was pointing out where he thought there were edges, and not seeing them at all myself. And then more data started coming in and they became clearer and clearer - though it may have been Isaac Asimov who said it first.

So my first reaction was more like “Huh, that looks really strange.” But I wouldn’t call myself disappointed. It was a puzzle that we needed to figure out.

FINKBEINER: Maybe befuddled is a better descriptor than disappointed.

MENG SU: I agree. We already knew of other bubble-like structures in the universe, but this was still quite a big shock. Finding these bubbles in the Milky Way wasn’t anticipated by any theories. When Doug first showed us the picture where you could start to see the bubbles, I immediately started to think about what could possibly produce this type of structure besides dark matter. I personally was less puzzled by the structure itself and more puzzled by how the Milky Way could have produced it.

SLATYER: But of course it’s also true that the structures we see in other galaxies have never been seen in gamma rays. As far as I know, beyond the question of whether the Milky Way could make a structure like this, there had never been an expectation that we would see a bright signal in gamma rays.

SU: That’s right. This discovery is still unique and, to me, punishing.

Fermi BubblesHints of the Fermi bubbles' edges were first observed in X-rays (blue) by ROSAT, which operated in the 1990s. The gamma rays mapped by the Fermi Gamma-ray Space Telescope (magenta) extend much farther from the galaxy's plane. (Credit: NASA's Goddard Space Flight Center)

TKF: Why were such bubbles not expected in the Milky Way, if they are seen in other galaxies?

FINKBEINER: It’s a good question. On the one hand we’re saying that these aren’t uncommon in other galaxies, while on the other hand we’re saying they were totally unexpected in the Milky Way. One of the reasons it was unexpected is that while every galaxy has a supermassive black hole at the center, in the Milky Way that black hole is about 4 million times the mass of the sun while in the galaxies in which we had previously observed bubbles, the black holes tend to be 100 or 1,000 times more massive than our black hole. And because we think it’s the black hole sucking in nearby matter that’s making most of these bubbles, you wouldn’t have expected a small black hole like the one we have in the Milky Way to be capable of this.

SU: For that reason, no one expected to see bubbles in our galaxy. We thought the black hole at the center of the Milky Way was a boring one that just sat there quietly. But more and more evidence is suggesting that it was very active a long time ago. It now seems that, in the past, our black hole could have been tens of millions of times more active than it is currently. Before the discovery of Fermi bubbles, people were discussing that possibility, but there was no single piece of evidence showing that our black hole could be that active. The Fermi bubble discovery changed the picture.

SLATYER: Exactly. Other galaxies that have similar looking structures are in fact quite different galactic environments. It’s not clear that bubbles we see in other galaxies with fairly similar shapes to the ones we see in the Milky Way are necessarily coming from the same physical processes. Due to the sensitivity of the instruments, we have no way to look at the gamma rays associated with these bubbles in other Milky Way-like galaxies – if they release gamma rays at all. The Fermi bubbles are really our first chance to look at anything like this close up and in gamma rays, and we just don’t know if many of the very puzzling features of the Fermi bubbles are present in other galaxies. It’s quite unclear at the moment the degree to which the Fermi bubbles are the same phenomenon as what we see in similarly shaped structures at other wavelengths in other galaxies.

SU: I think it’s actually very lucky that our galaxy has these structures. We get to look at them very clearly and with great sensitivity, allowing us to study them in detail.

SLATYER: Something like this could be present in other galaxies, and we would never know.

SU: Yes – and the opposite is true, too. It’s completely possible that the Fermi bubbles are from something we’ve never seen before.

FINKBEINER: Exactly. And, for example, the X-rays we do see coming from bubbles in other galaxies, those photons have a factor of a million times less energy than the gamma rays we see streaming from the Fermi bubbles. So we should not jump to conclusions that they come from the same physical processes.

SU: And, here in our own galaxy, I think more people are asking questions about the implications of the Milky Way’s black hole being so active. I think the picture and the questions are different now. Discovering this structure has very important implications to many key questions about the Milky Way, galaxy formation and black hole growth.

TKF: Doug and Meng, in a Scientific American article you coauthored with Dmitry Malyshev, you said that Fermi bubbles “promise to reveal deep secrets about the structure and history of our galaxy.” Will you tell us more about what type of secrets these might be?

SU: There are at least two key questions we’re trying to answer about the supermassive black holes in the center of each galaxy: How does the black hole itself form and grow? And, as the black hole grows, what’s the interaction between the black hole and the host galaxy?

I think that how the Milky Way fits into this big picture is still a mystery. We don’t know why the mass of the black hole in the center of the Milky Way is so small relative to other supermassive black holes, or how the interaction between this relatively small black hole and the Milky Way galaxy works. The bubbles provide a unique link for both how the black hole grew and how the energy injection from the black hole accretion process impacted the Milky Way as a whole.

FINKBEINER: Some of our colleagues at the Harvard–Smithsonian Center for Astrophysics conduct simulations where they can see how supernova explosions and black hole accretion events heat gas and drive it out of a galaxy. You can see in some of these simulations that things are going along just fine and stars are forming and the galaxy is rotating and everything is progressing, and then the black hole reaches some critical size. Suddenly, when more matter falls into the black hole, it makes such a big flash that it basically pushes most of the gas right out of the galaxy. After that, there’s no more star formation – you’re kind of done. That feedback process is key to galaxy formation.

SU: If the bubbles – like the ones we found – form episodically, that could help us understand how the energy outflow from the black hole changes the halo of the gas in the Milky Way dark matter halo. When this gas cools, the Milky Way forms stars. So the whole system will be changed because of the bubble story; the bubbles are closely linked to the history of our galaxy.

TKF: What additional experimental data or simulations are needed to really understand what’s going on with these bubbles?

SU: Right now, we’re focused on two things. First, from multi-wavelength observations, we’re looking to understand the current status of the bubbles – how fast they expand, how much energy is released through them, and how high-energy particles within the bubbles are accelerated either close to the black hole or inside the bubbles themselves. Those details we want to understand as much as possible through observations. Second, we want to understand the physics. For example, we want to understand just how the bubbles formed in the first place. Could a burst of star formation very close to the black hole help form the outflow that powers the bubbles? This can help us understand what kind of process forms these types of bubbles.

FINKBEINER: Any type of work that can give you the amount of energy released over specific timescales is really important to figuring out what’s going on.

SU: Truthfully, I think it’s amazing how many of the conclusions we drew from the very first observations of the bubbles still hold true today. The energy, the velocity, the age of the bubbles – all of these are consistent with today’s observations. All of the observations point to the same story, which allows us to ask more detailed questions.

TKF: That doesn’t often happen in astrophysics, where your initial observations are so spot-on.

FINKBEINER: This doesn’t always happen, it’s true. But we also weren’t very precise. Our paper says that the bubbles are somewhere between 1 and 10 million years old, and now we think they’re about 3 million years old, which is logarithmically right between 1 and 10 million. So, we’re pretty happy. But it’s not like we said it would be 3.76 million and were right.

TKF: What are the other remaining mysteries about these bubbles? What more do you hope to learn that we haven’t discussed already?

FINKBEINER: We have an age. I’m done. [laughter]

TKF: Ha! Now that does not sound like astrophysics.

SU: No, actually, we expect to learn many new things from future observations. We’ll have additional satellites launching in the coming years that will offer better measurements of the bubbles. One surprising thing we’ve found is that the bubbles have a high-energy cut off. Basically, the bubbles stop shining in high-energy gamma rays at a certain energy. Above that, we don’t see any gamma rays and we don’t know why. So we hope to take better measurements that can tell us why this cutoff is happening. This can be done with future gamma-ray energy satellites, including one called Dark Matter Particle Explorer that will launch later this year. Although the satellite is focused on looking for signatures of dark matter, it will also be able to detect these high-energy gamma rays, even higher than the Fermi Gamma-ray Space Telescope, the telescope we used to discover the Fermi bubbles. That's where the name of the structure came from.

Likewise, we’re also interested in the lower energy gamma rays. There are some limitations with the Fermi satellite we’re currently using - the spatial resolution is not nearly as good for low-energy gamma rays. So we hope to launch another satellite in the future that can view the bubbles in low-energy gamma rays. I’m actually part of a team proposing to build this satellite, and I'm glad to find a good name for it: PANGU. It’s still in the early stages, but hopefully we can get the data within 10 years. From this, we hope to learn more about the processes within the bubbles that lead to the emission of gamma rays. We need more data to understand this.

We’d also like to learn more about the bubbles in X-rays, which also hold key information. For example, X-rays could tell us how the bubbles affect the gas in the Milky Way’s halo. The bubbles presumably heat up the gas as they expand into the halo. We’d like to measure how much the energy from the bubbles is dumped into the gas halo. That’s key to understanding the black hole’s impact on star formation. A new German-Russian satellite called eRosita, planned to launch in 2016, could help with this. We hope its data will help us learn details about all the pieces of the bubble and how they interact with the gas around them.

FINKBEINER: I completely agree with what Meng just said. That’s going to be a very important data set.

SLATYER: Figuring out the exact origin of the bubbles is something I’m looking forward to. For example, if you make some basic assumptions, it looks like the gamma-ray signal has some very strange features. Particularly, the fact that the bubbles look so uniform all the way across is surprising. You wouldn’t expect the physics processes we think are taking place inside the bubbles to produce this uniformity. Are there multiple processes at work here? Does the radiation field within the bubbles look very different than what we expect? Is there an odd cancelation going on between the electron density and radiation field? These are just some of the questions we still have, questions that more observations – like the ones Meng was talking about – should shed light on.

FINKBEINER: In other words, we’re still looking in detail and saying, “That looks funny.”

Fermi Gamma-ray Space Telescope
The Fermi Gamma-ray Space Telescope collected the data that revealed the Fermi bubbles. (Credit: NASA's Goddard Space Flight Center)

TKF: It sounds like there are still many more observations that need to be made before we can fully understand the Fermi bubbles. But from what we do know already, is there anything that could fire up the galactic core again, causing it to create more such bubbles?

FINKBEINER: Well, if we’re right that the bubbles come from the black hole sucking up a lot of matter, just drop a bunch of gas on the black hole and you’ll see fireworks.

TKF: Is there a lot of matter near our black hole that could naturally set off these fireworks?

FINKBEINER: Oh sure! I don’t think it’ll happen in our lifetimes, but if you wait maybe 10 million years, I wouldn’t be surprised at all.

SU: There are smaller bits of matter, like a cloud of gas called G2 that people estimate has as much mass as perhaps three Earths, that will likely be pulled into the black hole in just a few years. That will probably not produce something like the Fermi bubbles, but it will tell us something about the environment around the black hole and the physics of this process. Those observations might help us learn how much mass it would have taken to create the Fermi bubbles and what types of physics played out in that process.

FINKBEINER: It’s true, we might learn something interesting from this G2 cloud. But this might be a bit of a red herring, since no reasonable model indicates it will produce gamma rays. It would take a gas cloud something like 100,000,000 times larger to produce a Fermi bubble.

SU: There’s a lot of evidence that the galactic center was a very different environment several million years ago. But it’s hard to deduce the overall story of exactly how things were in the past and what’s happened in the intervening time. I think the Fermi bubbles might provide a unique, direct piece of evidence that there was once much richer surrounding gas and dust that fed the central black hole than there is today.

Fermi dataData from the Fermi Telescope shows the bubbles (in red and yellow) against other sources of gamma rays. The plane of the galaxy (mostly black and white) stretches horizontally across the middle of the image, and the bubbles extend up and down from the center. (Credit: NASA's Goddard Space Flight Center)

TKF: The Fermi bubbles certainly remain an exciting area of research. So does dark matter, which is what you were originally looking for when you discovered the Fermi bubbles. How is that original dark matter hunt going?

FINKBEINER: We’ve really come full circle. If one of the most talked about types of theoretical dark matter particles, the Weakly Interacting Dark Matter Particle, or WIMP, exists, it should give off some sort of gamma-ray signal. It’s just a question of whether that signal is at a level that we can detect. So if you ever want to see this signal in the inner galaxy, you have to understand all the other things that make gamma rays. We thought we understood them all, and then along came the Fermi bubbles. Now we really need to thoroughly understand these bubbles before we can go back to looking for WIMPs in the center of the galaxy. Once we understand them well, we can confidently subtract the Fermi bubble gamma rays from the overall gamma-ray signal and look for any excess of gamma rays remaining that might come from dark matter.

Putting together quotations from Richard Feynman and Valentine Telegdi, “Yesterday’s sensation is today’s calibration is tomorrow's background.” The Fermi bubbles are certainly very interesting in their own right, and they’ll keep people busy for many years trying to figure out what they are. But they’re also a background or a foreground for any dark matter searches, and need to be understood for that reason too.

SLATYER: This is what I’m working on in my research these days. And the first question to what Doug just said is often, “Well, why don’t you just look for evidence of dark matter somewhere other than the inner galaxy?” But in WIMP models of dark matter, we expect the signals from the galactic center to be significantly brighter than anywhere else in the sky. So just giving up on the galactic center is not generally a good option.

Looking at the Fermi bubbles near the galactic center, we have found a promising signal that could potentially be associated with dark matter. It extends a significant distance from the galactic center, and has a lot of the properties that you would expect from a dark matter signal – including appearing outside the bubbles as well.

This is a very concrete case where studies of the Fermi bubbles uncovered something that may be related to dark matter – which is what we were looking for in the first place. It also emphasizes the importance of understanding what exactly is going on in the bubbles, so that we can get a better understanding of this very interesting region of the sky.

FINKBEINER: It would be a supreme irony if we found the Fermi bubbles while looking for dark matter and then while studying the Fermi bubbles we discovered dark matter.