The Fermi Bubbles: Two Giant Lobes Hiding Above Our Galaxy

We live inside the Milky Way and cannot see it as a whole — which is partly why it took until 2010 for astronomers to notice that two enormous structures were erupting from its center. The Fermi Bubbles, discovered by NASA's Fermi Gamma-ray Space Telescope, extend roughly 25,000 light-years both above and below the galactic plane, forming an hourglass shape roughly half the diameter of the galaxy itself. They are among the largest coherent structures in the Milky Way, and a decade after their discovery, the question of what produced them is still open.

What the Fermi Telescope Saw

The Fermi Gamma-ray Space Telescope, launched in June 2008, surveys the entire sky in gamma rays every three hours. Gamma rays — photons with energies above about 100 keV — are produced by the most energetic processes in the universe: particle collisions, relativistic jets, radioactive decay, and inverse Compton scattering where relativistic electrons boost low-energy photons to gamma-ray energies.

By 2010, Fermi's Large Area Telescope (LAT) had accumulated enough data to detect a previously unknown gamma-ray excess above and below the galactic center. Meng Su, Tracy Slatyer, and Douglas Finkbeiner at Harvard published the discovery paper in November 2010, describing two symmetric lobes with relatively uniform intensity and sharp edges extending roughly 50° (about 25,000 light-years, or 8 kiloparsecs) above and below the galactic center.

Fermi Bubbles — two large-scale gamma-ray and X-ray emitting structures extending approximately 25,000 light-years (8 kpc) north and south of the Milky Way's center, discovered in 2010 by the Fermi Gamma-ray Space Telescope's Large Area Telescope instrument.

Several features made the Fermi Bubbles immediately striking. The gamma-ray spectrum within the bubbles is harder (meaning relatively more high-energy photons) than the diffuse gamma-ray background from cosmic ray interactions in the galactic disk. The bubble edges are surprisingly sharp, suggesting they are bounded by a physical contact discontinuity — like the edge of an expanding shock front — rather than being a diffuse haze that gradually fades into the background. And they are symmetric to a high degree, implying the energy injection that created them came from close to the galaxy's dynamical center.

Source: Su, Slatyer & Finkbeiner, Astrophysical Journal (2010) — Discovery paper for the Fermi Bubbles, describing the morphology, spectrum, and symmetry of the gamma-ray lobes and their sharp edges.

The eROSITA Bubbles: The Larger Picture

Ten years after the Fermi discovery, a new space telescope provided a dramatic confirmation and extension of the picture. Germany's eROSITA (extended Roentgen Survey with an Imaging Telescope Array), launched in July 2019 aboard the Russian-German Spektr-RG spacecraft, conducted its first all-sky X-ray survey in 2020.

The results, published in Nature in December 2020, were striking. eROSITA detected two enormous X-ray-emitting bubbles — now called the eROSITA bubbles — extending about 80° above and below the galactic plane. In physical units, the eROSITA bubbles are roughly 45,000–50,000 light-years (about 14 kpc) tall, significantly larger than the Fermi Bubbles, though the two structures appear to be co-spatial at their inner boundaries.

The eROSITA bubbles emit soft X-rays (energies around 0.5–2 keV), consistent with hot gas at temperatures of about 3–4 million Kelvin. Their morphology suggests they are bounded by a shock or contact discontinuity propagating into the hot gas halo of the Milky Way. The total thermal energy contained in the eROSITA bubbles is estimated at roughly 10⁵⁶ ergs — enough energy to power the Sun for about 10 trillion years.

Source: Predehl et al., Nature (2020) — eROSITA first-light all-sky survey detection of the eROSITA bubbles, measuring their morphology, X-ray spectrum, and estimated thermal energy content.

The relationship between the Fermi and eROSITA bubbles is not entirely settled. The Fermi Bubbles are seen in gamma rays and represent the high-energy, non-thermal component of the structures. The eROSITA bubbles are seen in X-rays and represent hot thermal gas. Both structures share the same general axis of symmetry and similar spatial scale, which strongly suggests a common origin event, with the two observational signatures tracing different physical components of the same outflow.

Two Mechanisms That Could Explain Them

The central mystery is what energy source inflated these structures. The leading hypotheses fall into two broad categories.

Did Sagittarius A* Fire a Jet?

The most immediately appealing explanation is that the Milky Way's central supermassive black hole, Sagittarius A* (Sgr A*, with a mass of about 4 million solar masses), was far more active in the recent past than it is today. Currently Sgr A* is extraordinarily quiescent — one of the least luminous known supermassive black holes relative to its mass. But many lines of evidence suggest this has not always been the case.

In the AGN jet model, an active phase of Sgr A* several million years ago produced powerful relativistic jets perpendicular to the galactic plane. These jets deposited enormous amounts of energy into the galactic halo, driving a cocoon of hot gas outward and producing the sharp-edged bubbles now observed. The model predicts a bipolar, symmetric structure with energetics comparable to what is observed.

Supporting evidence includes the Magellanic Stream, a tidal stream of gas trailing the Magellanic Clouds that shows evidence of having been ionized by a powerful ultraviolet flash from the galactic center roughly 2–3 million years ago — consistent with an active Sgr A* at that time. Observations by the Hubble Space Telescope of the Magellanic Stream's ionization state, published in 2013 by Bland-Hawthorn and colleagues, were interpreted as evidence for a Seyfert flare from Sgr A* roughly 2–3 million years ago, which would have been 100,000 times more luminous than the current quiescent state.

Source: Bland-Hawthorn et al., Astrophysical Journal (2013) — HST spectroscopy of the Magellanic Stream showing elevated ionization consistent with a Seyfert-level flare from Sgr A* within the last ~2–3 million years.

Or Was It a Starburst Wind?

The alternative hypothesis invokes not the central black hole but the intense star-forming activity in the Central Molecular Zone — the dense ring of molecular gas surrounding Sgr A* within about 200 parsecs. This region hosts a cluster of massive young stars and has a star formation rate per unit volume far exceeding the rest of the galaxy.

In the starburst wind model, the collective effect of thousands of supernova explosions and stellar winds from the Central Molecular Zone drove a powerful galactic superwind over a period of tens of millions of years. This wind, accelerating gas away from the galactic plane, created the observed bubbles. The mechanism is directly analogous to the starburst-driven winds observed in nearby starburst galaxies like M82, which has prominent X-ray and optical emission extending kiloparsecs above its disk.

The starburst model has the advantage of invoking processes we can directly observe today (ongoing star formation near the galactic center) and mechanisms we see operating in other galaxies. Its challenge is that the sharp edges of the Fermi Bubbles are more naturally explained by the impulsive energy injection of a jet event than by the sustained, distributed energy input of a starburst wind.

Source: Crocker & Aharonian, Physical Review Letters (2011) — Starburst wind model for the Fermi Bubbles, arguing that the gamma-ray emission is consistent with hadronic cosmic ray interactions driven by galactic center star formation activity.

Why the Edges Are Sharp: Contact Discontinuities and Shocks

One of the most diagnostic features of the Fermi Bubbles — and one that constrains origin models significantly — is the sharpness of their edges. In typical diffuse emission from the galactic halo, intensity gradients are smooth and broad. The Fermi Bubble boundaries show intensity dropping by a factor of several within just a few degrees.

Contact discontinuity — a surface separating two fluid regions with different temperatures and densities but equal pressure, across which there is no bulk flow. Unlike a shock, there is no velocity jump across a contact discontinuity.

Shock front — a discontinuity in a supersonic flow where pressure, temperature, and density jump abruptly, propagating outward into undisturbed medium.

The sharp edges could be either contact discontinuities (if the bubble interior and exterior are in pressure equilibrium) or actual shocks (if the bubble is still expanding supersonically). X-ray observations with XMM-Newton and eROSITA suggest the presence of a weak shock at the eROSITA bubble boundary, with a Mach number of roughly 1.5–2, implying the structures are still slowly expanding.

A sharp contact discontinuity is more consistent with a past impulsive event — a jet or a brief intense flare — than with a steady ongoing wind. If energy is continuously injected, you expect a more gradual boundary. The impulsive nature of the boundary is one of the strongest arguments for an origin involving Sgr A* activity rather than (or in addition to) the starburst wind.

Source: Kataoka et al., Astrophysical Journal (2013) — Suzaku X-ray spectroscopy of the Fermi Bubble edges showing temperature enhancement consistent with a shock front and providing constraints on the origin of the structures.

What Is Inside the Bubbles?

The interior of the Fermi Bubbles is not empty. Gamma-ray emission fills the volume, implying the presence of either:

1. Relativistic electrons undergoing inverse Compton scattering off CMB photons or infrared photons from stars, boosting them to gamma-ray energies (leptonic model), or

2. Relativistic protons colliding with ambient gas and producing pions, which decay into gamma rays (hadronic model).

Distinguishing these two mechanisms is critical because they imply different energy inputs and different timescales. The leptonic model requires either very recent injection of electrons (because electrons lose energy quickly through synchrotron and inverse Compton radiation — on timescales of roughly 10 million years) or ongoing reacceleration inside the bubbles. The hadronic model allows older injection because protons lose energy much more slowly.

The current consensus, based on the hard spectral index and spatial uniformity of the gamma-ray emission, slightly favors a leptonic origin for the Fermi Bubble gamma rays. But the hadronic contribution may be significant and the question is not resolved.

Source: Ackermann et al. (Fermi-LAT Collaboration), Astrophysical Journal (2014) — Detailed spectral and morphological analysis of the Fermi Bubbles, constraining the gamma-ray emission mechanism and energy content.

Comparison: Fermi Bubbles vs. eROSITA Bubbles

Property	Fermi Bubbles	eROSITA Bubbles
Discovery year	2010	2020
Telescope	Fermi-LAT (gamma-ray)	eROSITA (X-ray)
Emission type	Gamma-ray (non-thermal)	Soft X-ray (thermal)
Height above/below plane	~25,000 ly (8 kpc) each	~45,000–50,000 ly (14 kpc) each
Energy content (estimated)	~10⁵⁵ ergs (non-thermal)	~10⁵⁶ ergs (thermal)
Physical interpretation	Relativistic particles	Hot gas ~3–4 million K
Edge sharpness	Very sharp	Sharp, possible weak shock

The eROSITA Bubbles: An Even Larger Structure Enveloping the Fermi Lobes

The Fermi Bubbles had been studied for a decade when the German-Russian Spektr-RG spacecraft, launched in July 2019, began its all-sky X-ray survey in 2020. The satellite's primary instrument, eROSITA (extended Roentgen Survey with an Imaging Telescope Array), scanned the sky in soft X-rays with sensitivity and angular resolution far exceeding any previous all-sky X-ray survey. The first results, published in Nature in December 2020 by Peter Predehl and colleagues at the Max-Planck-Institut für extraterrestrische Physik, revealed that the Fermi Bubbles are nested inside something considerably larger.

The eROSITA bubbles extend approximately 80° above and below the galactic plane. In physical units, this translates to roughly 45,000 light-years (about 14 kiloparsecs) above and below the galactic midplane — nearly double the height of the Fermi Bubbles, which extend about 25,000 light-years (8 kiloparsecs) in each direction. The eROSITA structures are co-spatial with the Fermi Bubbles at their inner boundaries, suggesting a common origin, but the X-ray emission reaches substantially farther out into the galactic halo.

Source: Predehl et al., Nature (2020) — "Detection of large-scale X-ray bubbles in the Milky Way halo," doi:10.1038/s41586-020-2979-0; eROSITA first all-sky survey detection of X-ray bubbles extending ~14 kpc above and below the galactic plane, co-spatial with the Fermi Bubbles at their inner boundaries.

The eROSITA bubbles emit soft X-rays in the 0.5–2 keV energy band, consistent with hot thermal gas at approximately 3 to 4 million Kelvin. They are bounded by a sharp edge interpreted as a weak shock propagating into the ambient hot gas halo of the Milky Way — the Mach number is estimated at roughly 1.5 to 2.0, meaning the outer boundary is expanding at about 1.5 to 2 times the local speed of sound in the halo gas, which is on the order of several hundred kilometers per second.

Thermal X-ray emission — radiation produced when hot plasma emits X-rays through bremsstrahlung (free-free) radiation and line emission from highly ionized atoms. Unlike the non-thermal gamma-ray emission of the Fermi Bubbles, which is produced by relativistic particles, the eROSITA bubble emission traces hot gas at temperatures of millions of Kelvin.

The energy budget of the eROSITA bubbles is enormous. Predehl et al. estimated the total thermal energy stored in the hot gas of the eROSITA bubbles at approximately 10⁵⁵ to 10⁵⁶ ergs. For reference, the Sun releases about 10³³ ergs per second; the eROSITA bubble energy content is equivalent to the Sun's total energy output over roughly 10 trillion years, or approximately 100,000 average supernova explosions. The Fermi Bubbles' non-thermal energy content is estimated at the lower end of this range, around 10⁵⁵ ergs. The two structures together represent one of the largest energy deposits in the Milky Way's recent history.

The relationship between the Fermi and eROSITA bubbles provides the strongest constraints yet on the timeline of Sgr A*'s last active period. If both structures trace the same outflow event — a past AGN outburst from the galactic center — the different sizes correspond to different propagation speeds. The Fermi Bubbles' hard gamma-ray emission traces relativistic particles, which travel close to the speed of light and can populate a larger or differently shaped volume than the slower-moving thermal gas traced by eROSITA. Alternatively, the eROSITA bubbles may represent an older, larger-scale outflow that pre-dates the event that created the Fermi Bubbles — implying at least two distinct energetic episodes in Sgr A*'s recent history, separated by millions of years.

A coordinated outburst model, in which both structures originate from a single sustained active phase of Sgr A* several million years ago, is currently the most widely discussed framework. Kinematic modeling suggests that if the eROSITA bubbles expanded at approximately 500 km/s from the galactic center, they would reach their current size in roughly 20 to 30 million years. The Fermi Bubbles, being smaller and bounded by sharper edges consistent with a more impulsive energy injection, may trace a younger outburst within the same extended active period. Combined age estimates from leptonic cooling timescales (how long relativistic electrons can emit gamma rays before losing their energy) and the dynamical expansion time bracket Sgr A*'s last significant active episode between approximately 2 and 15 million years ago — a geologically recent blink in the Milky Way's 13-billion-year history.

Source: Zhang & Guo, Astrophysical Journal (2020) — Numerical hydrodynamic simulations of the eROSITA and Fermi Bubbles as products of a common AGN jet event from Sgr A*, showing that a single outburst with energy ~10⁵⁵–10⁵⁶ ergs can reproduce the morphology of both structures.

The Microwave Haze and the S-PASS Connection

The Fermi Bubbles are not the only large-scale feature connected to this galactic center outflow. In microwave wavelengths, the WMAP satellite (and later Planck) detected a diffuse excess at the same position as the Fermi Bubbles, now called the microwave haze. This emission is consistent with synchrotron radiation from relativistic electrons spiraling in the galactic magnetic field — a natural complement to the inverse Compton gamma-ray emission inside the bubbles.

At radio frequencies, the S-PASS survey (S-band Polarization All Sky Survey) has mapped large-scale polarized radio emission associated with the bubble region, tracing the magnetic field structure within and around the lobes. The magnetic field appears to be swept up and compressed at the bubble edges, consistent with an expanding shock or contact discontinuity.

These multi-wavelength connections — gamma rays (Fermi), X-rays (eROSITA), microwaves (WMAP/Planck), radio (S-PASS) — mean the Fermi Bubbles are among the most thoroughly characterized large-scale structures in the Milky Way, even if their origin remains debated.

Source: Carretti et al., Nature (2013) — S-PASS radio survey detection of a large-scale polarized radio emission structure co-spatial with the Fermi Bubbles, indicating relativistic electron and magnetic field content.

Frequently Asked Questions

Q: Are the Fermi Bubbles dangerous to Earth or the solar system?
A: No. The Fermi Bubbles are centered on the galactic center, roughly 26,000 light-years from the solar system. Their gamma-ray emission, while energetically significant on galactic scales, produces a completely negligible flux at Earth's distance. The cosmic ray and gamma-ray backgrounds from the bubbles are orders of magnitude smaller than the ambient cosmic ray flux Earth already experiences.

Q: How do astronomers see structures above and below the galactic plane?
A: The galactic plane — the disk of the Milky Way — is opaque to optical light because of dust. But gamma rays and X-rays pass through dust without significant absorption, allowing Fermi and eROSITA to see structures at large distances from Earth including the galactic center region and the halo above and below it. This is the same reason X-ray and gamma-ray astronomy have revealed the galactic center environment at all.

Q: Is Sagittarius A still capable of producing a similar outburst?*
A: In principle, yes. Sgr A* has enough mass and surrounding gas to undergo another active phase if it accretes enough material. The current quiescent state — Sgr A* today emits at roughly 10⁻⁹ times its Eddington luminosity — is not permanent. Astronomers do not currently predict when a new active phase might begin; the triggering mechanism (accretion of a molecular cloud, a stellar disruption event) is stochastic. A recurrence on the scale needed to re-inflate the Fermi Bubbles would likely require millions of years of sustained elevated activity.

Q: What is the estimated age of the Fermi Bubbles?
A: Age estimates depend strongly on the assumed expansion velocity. If the bubbles expanded at roughly 1,000 km/s from the galactic center, the current size implies an age of about 6–9 million years. Estimates from different methods (leptonic cooling times, kinematic modeling, the Magellanic Stream ionization argument) cluster in the range of 2–15 million years. The range reflects genuine physical uncertainty rather than observational error — the bubbles' dynamics are complex.

Q: Why were the Fermi Bubbles missed before 2010?
A: The galactic center region is extremely bright in gamma rays from conventional sources — cosmic ray interactions with gas, supernovae, pulsars. Separating the bubble emission from this background required careful subtraction using models of the diffuse galactic emission. Earlier gamma-ray telescopes lacked either the sensitivity or the sky coverage to see the bubbles above this foreground. Fermi-LAT's combination of full-sky coverage, angular resolution, and multi-year integration time made the detection possible.

Q: Could the Fermi Bubbles affect the Magellanic Clouds?
A: The Large Magellanic Cloud sits at about 50 kiloparsecs from the galactic center, well outside the current extent of even the eROSITA bubbles. However, if the galaxy underwent a more energetic AGN phase in the past — the Seyfert flare hypothesis inferred from Magellanic Stream ionization — the radiation and wind from that event could have interacted with the LMC and its surrounding gas. Ongoing observations of the Magellanic Stream's ionization structure are probing this possibility.

The Takeaway

The Fermi Bubbles and their larger X-ray counterparts, the eROSITA bubbles, are among the most spectacular large-scale structures in the Milky Way — and we had no idea they existed until 2010. They represent an enormous injection of energy into the galactic halo, totaling at least 10⁵⁶ ergs, at some point in the past few million to tens of millions of years. Whether that energy came from a jet powered by Sagittarius A* during an active phase, from a galactic-center starburst wind, or from some combination of both remains the central open question. The answer will reveal something important about the history of our own galaxy — the kind of energetic events that shaped its halo and potentially affected its star formation history — and about how galaxies in general regulate the cycling of gas between their disks and their surrounding halos.

---

Read next:

• Sagittarius A*: The Milky Way's Central Black Hole
• Black Holes Explained: How They Work
• Gravitational Waves Explained: Listening to the Universe