Solar Flare

What a flare actually is

A solar flare is a sudden, intense brightening of a localised region of the Sun's atmosphere, caused by the rapid conversion of stored magnetic energy through magnetic reconnection. The brightening extends across the electromagnetic spectrum, from gamma rays at the impulsive peak through hard X-rays at the chromospheric footpoints, soft X-rays from the heated coronal loop top, EUV and UV in the post-flare arcade, Hα and white-light at the chromosphere and photosphere, and radio bursts from accelerated electrons. The whole event runs from a few minutes (the impulsive phase) through several hours (the gradual decay).

The energy comes from the magnetic field in a solar active region — the bipolar magnetic-flux concentrations that house sunspots. Photospheric convection slowly twists and shears the field lines anchored at sunspot footpoints, building up free magnetic energy (above the potential, current-free configuration). When the topology becomes unstable — typically after the appearance of an X-line, a region where oppositely directed field lines can reconnect — the system snaps. Field lines reconfigure, plasma is accelerated and heated, and the released energy radiates and propagates outward.

The reconnection mechanism

Magnetic reconnection is the central physics. In ideal MHD, magnetic field lines are frozen into the plasma and cannot change topology. In real plasma there is finite resistivity (η ~ 10⁻⁵ Ω·m for the corona, but enhanced enormously in thin current sheets by collisionless plasma effects). When two regions of oppositely directed field are pushed together, a thin current sheet forms; in the sheet, η dominates and field lines can "break" and "reconnect" into new configurations.

The standard model of an eruptive flare — the CSHKP model (Carmichael 1964, Sturrock 1966, Hirayama 1974, Kopp-Pneuman 1976) — has three components:

• Pre-flare phase. A flux rope or sigmoidal arcade slowly destabilises over hours. Heliospheric current sheets and X-lines develop.
• Impulsive phase. Reconnection at the X-line accelerates particles to MeV–GeV energies. Hard X-rays and microwaves peak; energetic particles stream down along field lines and slam into the chromosphere at the footpoints.
• Gradual phase. Heated chromospheric plasma evaporates upward, filling the post-flare loop arcade. Soft X-ray and EUV emission from the heated arcade decays over hours.

The GOES X-ray classification

Since 1976, the GOES satellite series has continuously monitored solar soft X-rays in the 1-8 Ångström band. The peak flux during a flare gives its GOES class — a logarithmic letter scale (each letter is a factor of 10) with a 1-9.9 sub-numerical:

Class	Peak X-ray flux (W/m², 1-8 Å)	Typical energy	Frequency at solar max	Earth impact
A	< 10⁻⁷	10²⁰ J	Many per day	None detectable
B	10⁻⁷ – 10⁻⁶	10²¹ J	Many per day	None detectable
C	10⁻⁶ – 10⁻⁵	10²² J	10–20 per day	Minor HF radio fade
M	10⁻⁵ – 10⁻⁴	10²³ J	1–3 per day	HF radio fade, minor satellite anomaly
X	≥ 10⁻⁴	10²⁴ – 10²⁵ J	~10 per year (max)	Wide-area HF blackout, possible CME
X10+	≥ 10⁻³	10²⁵ J	1–3 per cycle	GPS errors, satellite damage
X20+	≥ 2 × 10⁻³	3 × 10²⁵ J	~1 per several decades	Power grid GIC, severe satellite impact

Five letter classes, two decades of sub-numerical resolution. Below A1 the flux is dominated by the quiet-Sun X-ray background. The largest historical events (Carrington 1859, ~X45; Halloween 2003, X28; Bastille Day 2000, X5.7) populate the extreme tail.

Worked example: energy of an X-class flare

Compute the total radiated energy of an X10 flare with peak flux F_peak = 10⁻³ W/m² in 1-8 Å, observed from 1 AU. Assume the peak duration is 5 minutes (300 s) and the soft X-ray band carries ~ 1% of the total flare energy (the rest is in EUV, hard X-ray, particles, kinetic energy of CME):

Solid-angle correction: assume isotropic emission, distance d = 1 AU = 1.496 × 10¹¹ m.

P_X    = F_peak × 4π d²
        = 10⁻³ × 4π × (1.496 × 10¹¹)²
        = 10⁻³ × 2.81 × 10²³
        = 2.81 × 10²⁰ W

Energy in 1-8 Å during 5-min peak (rough rectangle):
E_X    ≈ P_X × Δt
        = 2.81 × 10²⁰ × 300
        = 8.4 × 10²² J

Total flare energy (X-rays = 1% of total):
E_total = E_X × 100
        = 8.4 × 10²⁴ J
        ≈ 10²⁵ J.

An X10 flare therefore releases on the order of 10²⁵ J — about 10⁷ times the largest hydrogen-bomb test, but only ~ 1% of the Sun's total power output for the few minutes it lasts. The same energy in equivalent TNT is ~ 2 × 10¹⁵ tonnes; equivalent to detonating 50 million Tsar Bombas simultaneously. The largest recorded events (Carrington-class) reach ~ 5 × 10²⁵ J.

Famous solar flares

• Carrington Event (1-2 September 1859). Estimated X45, the largest in recorded history. Observed by Richard Carrington at Redhill, Surrey, as a white-light flare in active region 520. Auroras visible as far south as the Caribbean, Hawaii, and central Mexico. Telegraph systems failed across Europe and North America; some operators reported sparks and shocks from their equipment. A modern-equivalent event would cause trillion-dollar damage to power grids, satellites, and HF communications.
• 1989 Quebec Storm (9-13 March 1989). X15 flare on 9 March 1989, followed by major CME impacts. Geomagnetically induced currents tripped 21,000 km of Hydro-Québec transmission lines; the entire province blacked out for 9 hours.
• Bastille Day Flare (14 July 2000). X5.7 with associated high-energy proton event. Damaged solar arrays on satellites; produced the strongest proton event observable from Earth (10⁴ pfu > 10 MeV).
• Halloween Storms (29 October-4 November 2003). Sequence including X17.2 (28 October) and X28 (4 November) — the largest GOES-measured flare on record. ADEOS-2 satellite lost. Air traffic re-routed away from polar routes. Astronaut crew on ISS took shelter in shielded area.
• September 2017 Storm (6 September 2017). X9.3 — the largest of solar cycle 24. Disrupted HF radio across the Caribbean during Hurricane Irma recovery operations; widespread aurora observations.
• Gannon Storm (May 2024). Sequence of X-class flares (X5.8 on 11 May, X8.7 on 14 May) from active region 13664. Auroras observed at 25° latitude, far below historical norms. Largest sequence since Halloween 2003.
• October 2024 X9.0 (3 October 2024). Single event of cycle 25 (peaked late 2024).

Variants and related phenomena

• Confined flare. A flare with no associated CME — about 40% of M and X events. Reconnection releases magnetic energy locally but the overlying field is strong enough to confine the disturbance.
• Eruptive flare. A flare accompanied by a CME — about 60% of M and X events. The flux rope erupts and the CME carries plasma and magnetic field outward through the heliosphere.
• Two-ribbon flare. Standard CSHKP geometry — two parallel ribbons of brightening in the chromosphere, marking the footpoints of the post-flare loop arcade.
• Compact (impulsive) flare. Single-loop flare with no observable arcade; reconnection within a single magnetic structure.
• White-light flare. A flare so energetic that the photosphere itself brightens visibly in continuum. Carrington's 1859 observation was a WLF. About 10% of X-class events produce detectable white-light continuum.
• Stellar flares ("superflares"). On other stars, especially M dwarfs and active solar-type stars, flares 10² – 10⁴ times larger than Carrington are observed. Kepler has catalogued thousands. The risk of a Carrington-equivalent or larger superflare on the modern Sun is estimated at ~ 1% per century, with substantial uncertainty.

Where solar flares show up

• Space weather. NOAA Space Weather Prediction Center issues alerts; FAA reroutes aurora-region flights; satellite operators put spacecraft in "safe mode"; HF radio operators experience radio fades. Carrington-class events are tier-1 catastrophe risks for power grids.
• GPS / navigation. Ionospheric disturbances from X-ray and EUV from flares delay GPS signals; reception in mid-latitude users can degrade by metres during an M+ event.
• Astronaut radiation. Solar energetic particles from flare-CME pairs deliver radiation doses to astronauts; ISS crew shelter in shielded areas during major events; lunar/Mars missions need radiation-shielded refuges.
• Aurora. Flare-driven CMEs that impact Earth produce geomagnetic storms and aurorae; extreme events drive auroras to mid-latitudes (e.g., 2024 May Gannon Storm at 25° N).
• Stellar astrophysics. Solar flares are the only flare type we can directly resolve. They serve as the calibration for stellar flare physics on cooler dwarfs (proxy for habitability research) and for activity-induced variability in young exoplanet hosts.
• Beryllium-10 / Carbon-14 records. Massive solar particle events leave isotopic spikes in ice cores and tree rings. The Miyake events (774-775 AD, 993-994 AD, 660 BCE) record ancient extreme-event analogs.

Common pitfalls

• Conflating flares with CMEs. A flare is a localised radiation event; a CME is a plasma eruption. They are correlated (60% of large flares have a CME) but distinct in physics and impact. The CME, not the flare X-rays, drives geomagnetic storms days later.
• Assuming flare X-rays reach Earth's surface. Soft X-rays are stopped by the atmosphere above ~ 100 km. Ionospheric disturbances are felt on the ground (HF radio); direct X-ray exposure is not.
• Reading the GOES class as total energy. The class is the peak soft X-ray flux in one band. Total radiated energy spans many bands and is roughly proportional to the class but with order-unity scatter.
• Treating reconnection as fully understood. The basic picture is robust, but the precise mechanisms by which reconnection accelerates particles to GeV energies and how energy is partitioned among plasma heating, particle acceleration, and bulk flow remain active research.
• Forgetting the eruptive-vs-confined distinction. Confined flares (no CME) and eruptive flares (with CME) have very different geomagnetic impact even at the same GOES class.