Sunquakes: How a Solar Flare Rings the Sun Like a Bell

What a Sunquake Is: Striking the Sun's Bell

A sunquake is a packet of acoustic (sound) waves in the Sun's interior that is set ringing impulsively by a solar flare. The Sun is always full of sound — roughly ten million overlapping acoustic p-modes that convection excites continuously, the subject of helioseismology. A sunquake is different: instead of a steady hum stirred up from below, it is a single sharp strike from above, so it produces a clean, localized transient rather than a global standing wave.

What we actually see is a set of expanding circular ripples in the line-of-sight velocity (Doppler) map of the photosphere, emanating from the flare's footpoint. The analogy to a pebble dropped in a pond is close to literal — but with two twists that make sunquakes strange and diagnostic. First, the waves are not surface ripples on water; they are sound waves that dive into the star and refract back up. Second, because of that diving path, the rings appear to speed up as they move outward, the opposite of ripples fading and slowing on a pond. Watching a sunquake is therefore watching the Sun's interior reveal itself through a single, well-timed blow.

The Trigger: From Reconnection to a Mechanical Punch

The engine is the flare itself. A solar flare releases ~10³²–10³³ erg when stressed coronal magnetic fields undergo magnetic reconnection, converting stored magnetic energy into heat, radiation, bulk flows, and beams of accelerated particles. The question a sunquake poses is: how does energy released high in the tenuous corona deliver a mechanical impulse to the dense photosphere millions of times denser, thousands of kilometers below?

The leading chain of events runs like this:

• Energy transport downward. Reconnection accelerates electrons (and, in some events, protons) that stream down the magnetic loops, plus a front of thermal conduction. These deposit their energy in the chromosphere in seconds — the flare's brief, violent impulsive phase, marked by a hard X-ray spike.
• Explosive evaporation and condensation. The chromosphere is flash-heated to ~10⁷ K and over-pressurized. Most of it blows upward as chromospheric evaporation, filling the flare loops with the hot plasma that glows in soft X-rays. By momentum conservation, a thin, cool, dense layer is driven downward — a chromospheric condensation moving at tens of km/s.
• The blow lands. That downward-plowing condensation (plus, possibly, a jerk from the reconfiguring magnetic field — see the driver debate below) delivers a sudden overpressure to the photosphere over a minute or two. That is the ‘pebble.’ It launches an acoustic wave straight down into the interior.

Crucially, the strike is coincident in time and place with the flare's impulsive phase and often with the flare's white-light kernels and hard X-ray footpoints — a key clue that whatever excites the sunquake also produces the deepest, most concentrated heating.

Why the Ripples Appear to Accelerate

The most counterintuitive feature — the outward acceleration — is pure geometry of a refracting medium. The sound speed inside the Sun scales as c ≈ √(γP/ρ) ≈ √(γk_BT/μm_H), and since temperature climbs steeply with depth (from ~5,800 K at the surface toward millions of kelvin inside), sound travels faster the deeper you go. A wave launched downward at an angle therefore does not travel in a straight line: its deeper part outruns its shallower part, so the ray bends back toward the surface, resurfacing some distance away like a stone skipped across a pond.

Now consider the whole spray of rays from the strike. Rays launched at steeper angles dive deeper, reach faster layers, and come back up farther out and later. So the ripple that surfaces at large radius is intrinsically the fast, deep-diving one, while the near-in ripple sampled only the slow shallow layers. Plotted as a wavefront, the ring's apparent horizontal speed grows with distance — it looks like acceleration, though no single wave is speeding up. In the discovery event this apparent speed rose from ~10 km/s near the source to about 100 km/s at the outer edge, and the wave stayed recognizable out to roughly 120,000 km — about ten Earth diameters — over the course of an hour before fading into the granulation. On a time–distance diagram the sunquake traces a curved ridge, the same ridge shape that encodes the interior sound-speed profile in helioseismology.

The Numbers: Modest Energy, and Why Big Flares Often Stay Silent

By flare standards a sunquake is energetically cheap. Estimates of the seismic energy radiated into the interior span roughly 10²⁷–10²⁹ erg (~10²⁰–10²² J) — only about 10⁻³ to 10⁻⁵ of the flare's total energy budget. That is still enormous compared with terrestrial seismology: the 1996 sunquake was described as carrying tens of thousands of times the seismic energy of the great 1906 San Francisco earthquake. The point of measuring it is not the raw number but the efficiency — what fraction of a flare's energy actually reaches the photosphere as a coherent mechanical impulse.

The most surprising empirical fact is that sunquake strength correlates only weakly with flare class. Plenty of powerful X-class flares produce no detectable quake, while some comparatively modest flares ring the Sun loudly. Whether a flare is ‘acoustically active’ seems to depend less on total energy than on:

• How impulsively the energy is dumped — a fast, sharp deposit couples to acoustic waves far better than a slow one.
• How deep the heating and momentum reach — the driver must load the dense photosphere, not just the tenuous upper chromosphere.
• Magnetic geometry — compact, low-lying, strongly sheared footpoints appear to favor quakes, and the resulting ripples are frequently anisotropic (stronger in one direction), hinting at a directional driver rather than an isotropic explosion.

How We See Them: Doppler Ripples and Acoustic Holography

Sunquakes are detected the same way p-modes are: through periodic Doppler shifts of a photospheric spectral line, which measure the surface moving toward and away from us. The workhorses are space-based, above the blurring atmosphere and free of day–night gaps: SOHO's Michelson Doppler Imager (MDI, 1996–2011), which caught the first sunquake, and its far sharper successor, SDO's Helioseismic and Magnetic Imager (HMI, 2010–), which returns full-disk Doppler images every 45 seconds. Two complementary techniques turn those data into a sunquake:

• Time–distance imaging. The original method of Alexander Kosovichev and Valentina Zharkova: subtract a running background from the Doppler movie and watch the expanding rings directly, then read off the curved time–distance ridge. This works only for the strongest, cleanest events, where the ripple rises above the ever-present convective noise.
• Helioseismic (acoustic) holography. Pioneered by Charles Lindsey and Douglas Braun (and applied to the 1996 event by Donea, Braun & Lindsey in 1999), this back-propagates the observed surface wavefield to a chosen depth and time to compute the egression power — effectively focusing the Sun's own sound to pinpoint a compact subsurface seismic source. It is far more sensitive than eyeballing ripples and can flag ‘acoustically active’ flares whose surface waves are hidden in noise.

HMI's uninterrupted cadence has been transformative: sunquakes went from a handful of curiosities to a growing catalog, with well-studied events including the 15 February 2011 X2.2 flare (the first X-flare of Solar Cycle 24), the 10 September 2014 X1.6, and the powerful 6 September 2017 X9.3 flare.

Sunquakes vs the Flare's Other Waves — and the Driver Debate

A single flare launches a whole family of waves, and it is easy to confuse them. High in the corona a fast-mode magnetosonic EUV (‘EIT’) wave ripples outward at hundreds of km/s; its skirt sweeping the chromosphere is seen as a Moreton wave at ~1000 km/s; a coronal mass ejection may erupt above. The sunquake is the only member that travels into the star, as sound, at apparent speeds of tens of km/s. Historically the chromospheric Moreton wave and the sub-surface sunquake were suspected to be two faces of the same flare blast wave, and in some events they do appear linked.

What actually delivers the mechanical punch remains an open question, with three main candidates that may all contribute in different events:

• Hydrodynamic / backwarming shock. The downward chromospheric condensation and intense radiative heating of the deep atmosphere drive a pressure transient into the photosphere (Donea, Lindsey and collaborators).
• Magnetic back-reaction (the ‘magnetic jerk’). As the coronal field reconnects and ‘implodes’ to a lower-energy state (Hudson's magnetic-implosion picture), the photospheric field turns abruptly more horizontal, and the sudden change in Lorentz (Maxwell-stress) force jerks the surface downward — a mechanism quantified by G. Fisher and colleagues (2012), supported by observed permanent steps in the photospheric magnetic field at flare footpoints.
• Particle-beam momentum. Direct impact of flare-accelerated protons penetrating deep, favored by V. Zharkova and collaborators and consistent with sunquakes coinciding with hard X-ray/gamma-ray footpoints.

Beyond the driver, the deep puzzles are why the energy–flare-class correlation is so weak, why many ripples are directional, and whether ‘starquakes’ ring other stars during their far larger superflares — a signal we have not yet been able to isolate. Each new HMI event is a fresh, well-timed hammer-blow on the Sun, and helioseismologists are still learning to read the ring.

One flare, several waves: a sunquake is only the sub-surface member of a family of flare-driven disturbances.

Disturbance	Layer / medium	Restoring force / nature	Typical speed
Sunquake (seismic wave)	Sub-photosphere / interior	Acoustic (pressure) — refracted p-modes	Apparent ~10–100+ km/s (accelerating outward)
Moreton wave	Chromosphere	Skirt of a coronal MHD fast-mode shock sweeping the surface	~500–1500 km/s
EUV ('EIT') wave	Low corona	Fast-mode magnetosonic wave / disturbance front	~200–500 km/s
Coronal mass ejection	Corona to interplanetary space	Bulk ejection of magnetized plasma	~300–3000 km/s
Ordinary p-mode oscillations	Interior (global)	Acoustic, stochastically driven by convection	Standing modes, ~5-minute period