Coronal Heating Problem

The temperature profile that shouldn't exist

Walk outward from the Sun's center. The core, where deuterium fuses to helium, is about 15 million K. Heat diffuses outward through the radiative zone — photons random-walking off electrons for around 170,000 years — and through the convective zone, where buoyancy plumes carry heat the final 200,000 km. By the time energy reaches the photosphere, the temperature has fallen monotonically to about 5778 K. This is the visible surface, the layer where the Sun becomes transparent and photons escape to space. So far so thermodynamic.

Now keep going outward. Across the 2000-km-thick chromosphere the temperature climbs from about 4000 K at the temperature minimum to ~25000 K at the top. Across a few thousand more km — the transition region — it explodes from 25000 K to over 10⁶ K. The corona itself, the diffuse outer atmosphere extending millions of km into space, sits at 1–3 million K. So the temperature profile is non-monotonic: it falls, then rises, then rises again, ending more than 200 times hotter than the surface immediately below.

This is not a small effect. A doubling of temperature is unusual; a 200× inversion across 5000 km is the second law of thermodynamics shouting at the observer. Heat cannot flow up this gradient by ordinary radiation or conduction. Some non-thermal mechanism must transport energy through the photosphere, past the chromosphere, and dissipate it in the corona. Identifying that mechanism is the coronal heating problem.

History of the puzzle

In the 1860s, observers of solar eclipses noted a bright green emission line in the corona at 5303 Å. It didn't match any known terrestrial element, so a new element was proposed — "coronium" — by analogy to "helium" (named for the Sun before being found on Earth in 1895). This was popular until 1939, when Walter Grotrian showed that high ionization states of common elements could plausibly emit at coronal wavelengths if the temperature were extremely high. Bengt Edlén closed the case in 1942 by identifying the green line specifically as Fe XIV — iron with 13 electrons stripped off — which requires a plasma temperature of around 2 million K. Coronium didn't exist; the corona was just very hot iron.

This identification immediately raised the puzzle. The photosphere is 5778 K. Iron in the photosphere is at most singly ionized. The corona, sitting above the photosphere, is 2 million K with iron stripped of half its electrons. The question of how this happens has driven 80 years of theory and observation.

Energy budget — how much heating is required?

You can estimate the energy required from the corona's radiative losses plus its conductive losses to the chromosphere plus the kinetic energy carried out by the solar wind. The standard accounting from Withbroe and Noyes (1977) gives, in steady state:

Region          Heating flux required
─────────────── ──────────────────────
Coronal hole    ≈ 800   W/m²
Quiet corona    ≈ 300   W/m²
Active region   ≈ 10⁴   W/m²

For comparison, the photospheric kinetic energy flux from convection is ~10⁸ W/m², so only a tiny fraction (10⁻⁵–10⁻³) of photospheric mechanical energy needs to be converted to corona heating. The mechanism can be wildly inefficient and still work in principle. The challenge is the transport and dissipation chain: how energy gets out of the photosphere into a usable form, then up through the chromosphere, then thermalized in the corona.

Mechanism 1: Nanoflares (DC heating)

Eugene Parker's 1988 paper "Nanoflares and the Solar X-ray Corona" proposed that the corona is heated by an enormous number of tiny reconnection events. The picture is this: photospheric convection drags the footpoints of coronal magnetic field lines around randomly. The field above is anchored at both ends, so it gets braided — like braiding hair by walking the ends in circles. Parker showed that a braided field with mean angle of misalignment θ between adjacent strands has a free magnetic energy density per unit length of roughly B²·tan²(θ)/(8π). Once θ exceeds about 14°, a current sheet forms, becomes resistively unstable, and reconnects in a "nanoflare."

The total energy of a single nanoflare is around 10¹⁷ joules — one-billionth of an X-class flare, or about 10⁻⁹ of a typical solar flare. Parker predicted that there should be approximately 10⁷ nanoflares per second over the whole Sun. Their integrated heating power matches the required ~10⁻³ of photospheric mechanical energy.

The signatures predicted by the nanoflare model are: (a) a power-law size distribution dN/dE ∝ E^(-α) with α between 2 and 3, where α > 2 means small events dominate the integrated energy; (b) extremely hot, transient ~10 MK plasma above the bulk coronal temperature of 1–3 MK; and (c) characteristic time profiles of EUV intensity that look impulsive rather than steady. All three signatures have been searched for in Hinode, SDO, and IRIS data. The evidence is mixed: power-law indices closer to 1.7 (below the critical α = 2) have been measured, suggesting nanoflares contribute but don't dominate. Hot plasma above 5 MK has been detected in active regions, supporting impulsive heating there, but is rare in the quiet Sun.

Mechanism 2: Alfvén waves (AC heating)

Hannes Alfvén showed in 1942 that magnetic field lines under tension transmit transverse waves with characteristic speed v_A = B/√(μ₀ρ). In the corona, with B ≈ 10⁻³ T and ρ ≈ 10⁻¹² kg/m³, the Alfvén speed is about 1000 km/s. Photospheric convection — granules turning over every 8 minutes — jostles magnetic footpoints. Any vertical motion launches an Alfvén wave up the field line, carrying energy density ρv_perp² and propagating energy flux ρv_perp²·v_A.

Direct measurements of Alfvénic motions in the chromosphere and corona by CoMP (the Coronal Multi-Channel Polarimeter on Sacramento Peak) and IRIS find Poynting fluxes of 10²–10⁴ W/m² — sufficient to power coronal heating in principle. The catch is dissipation. Alfvén waves are linearly non-dissipative; in a uniform, ideal MHD plasma they propagate forever. To deposit their energy, they need a dissipation mechanism: phase mixing (waves on adjacent field lines with different v_A get out of phase and develop strong gradients), resonant absorption (waves matching a local Alfvén resonance get absorbed), or turbulent cascade (nonlinear wave-wave interaction produces small-scale structure that dissipates resistively or by ion-cyclotron damping).

Recent Parker Solar Probe data show strong Alfvénic turbulence in the open-field regions feeding the fast solar wind, consistent with wave-driven heating there. They also show "switchbacks" — abrupt S-shaped magnetic field reversals — that may be the imprint of small-scale reconnection events at the corona's base. Both mechanisms operate, in the same regions, at the same time.

Worked example: nanoflare energy budget

Can a population of 10¹⁷ J nanoflares plausibly heat a quiet-Sun coronal region requiring 300 W/m²?

Required power per m² of surface  = 300 W/m²
Sun's surface area                = 4π·(6.96×10⁸ m)² = 6.09×10¹⁸ m²
Total quiet-Sun corona heating    = 300 × 6.09×10¹⁸  = 1.83×10²¹ W

Per-nanoflare energy             = 10¹⁷ J
Nanoflare rate to balance budget = 1.83×10²¹ / 10¹⁷  = 1.83×10⁴ /s
Distributed over the surface     = 1.83×10⁴ / 6.09×10¹⁸  = 3.0×10⁻¹⁵ /m²/s

So you need about 18,000 nanoflares per second across the whole Sun, or one nanoflare per square kilometer of solar surface every 1000 years. Per unit photospheric area this is well below current observation thresholds — Hinode/EIS can resolve individual events at the picoflare level (10¹⁴ J) but cannot integrate that statistically over the full corona. The number is plausible. What is harder to test is whether the distribution of nanoflare energies has the right power-law slope α > 2 to make small events dominate, or whether you need larger "microflares" doing most of the work. The literature is split.

Why this has resisted solution for 80 years

Three reasons. First, both candidate mechanisms predict the right total power, so calorimetry alone can't distinguish them. Second, their characteristic spatial and temporal scales — sub-arcsecond and sub-second — sit at or below the resolution limits of current instruments. Third, both mechanisms can co-occur — and probably do — so an experiment looking for nanoflares may see a contribution while missing the dominant wave signature, and vice versa. The reality may be a regime-dependent mix: nanoflare-dominated in closed-field active regions where braiding is strong, wave-dominated in open-field coronal holes feeding the fast solar wind.

Where the candidate mechanisms differ

Property	Nanoflares (DC heating)	Alfvén waves (AC heating)	How to tell apart
Energy storage	Magnetic free energy in braided coronal field	Kinetic energy of photospheric motions launched as waves	Difficult — both end up as magnetic perturbation
Dissipation timescale	Impulsive — seconds to minutes per event	Continuous — sustained wave damping	High-cadence EUV intensity statistics
Predicted plasma signature	Brief 5–10 MK plasma populations	Steady 1–3 MK with Alfvén-wave-broadened lines	Differential emission measure tail above 5 MK
Frequency-energy spectrum	Power law dN/dE ∝ E^(-α), α ≈ 2.5 (Parker 1988)	Continuous turbulent cascade	Slope α — α > 2 favors nanoflares
Preferred environment	Closed-field active regions with footpoint braiding	Open-field coronal holes feeding solar wind	Wavelet analysis of CoMP/IRIS spectra
Driving timescale	Minutes to hours (granulation turnover × build-up)	5-minute photospheric oscillations to seconds	Fourier analysis of footpoint motions
Key prediction signature	Hot, dense, brief loop brightenings	Smooth Alfvénic line broadening proportional to height	Combined imaging + spectroscopy at high cadence

Variants and related mechanisms

• Type-II spicules. Thin chromospheric jets identified by De Pontieu et al. (2007) on Hinode/SOT data. Faster (50–150 km/s) and shorter-lived than classical spicules; sometimes contain hot plasma at 10⁶ K. Proposed as a separate mass-and-energy supply to the corona, distinct from waves and reconnection — though the heating mechanism inside the spicule is itself debated.
• Resonant absorption. Specific Alfvén-wave dissipation mode where the wave frequency matches the local Alfvén frequency on a thin shell of field lines, causing strong shear amplification and dissipation. Quantitatively viable in coronal-loop models (Ruderman & Roberts 2002).
• Phase mixing. Adjacent field lines with different B and ρ have different v_A; waves launched in phase get out of phase as they propagate, creating cross-field gradients that dissipate by Ohmic resistivity. Heyvaerts & Priest 1983 is the canonical reference.
• Turbulent reconnection. Lazarian & Vishniac 1999 — large-scale reconnection in a turbulent MHD plasma proceeds much faster than classical Sweet-Parker reconnection, removing one of the bottlenecks of the nanoflare scenario.
• Ion-cyclotron damping. High-frequency Alfvén waves can damp via cyclotron resonance with heavy ions, producing the strong perpendicular ion temperatures seen in coronal hole spectra. UVCS/SOHO observations of O VI line widths support this.

Common misconceptions

• "It's solved." No. Both candidate mechanisms operate; their relative importance is contested. Operational space-weather models still use empirical heating parameterizations rather than first-principles physics.
• "Heat just leaks up from the core." The core temperature is 15 million K, but the radiative diffusion timescale to the surface is 170,000 years. The corona's heating is essentially decoupled from the core; it comes from local magnetic energy in the convection zone and above.
• "Magnetic reconnection violates conservation laws." It does not. Reconnection converts magnetic energy to kinetic and thermal energy of the plasma; total energy and momentum are conserved. What changes is the topology of the field, freeing energy stored in braided magnetic configurations.
• "The corona is hot because the Sun's surface is hot." The surface is 200× cooler than the corona. The heat flow is bottom-up only because of the non-thermal energy carrier (waves or reconnection), not because of ordinary conduction.
• "All stars must have the same mechanism." The mechanism scales with stellar magnetic activity, which scales with rotation. Slow rotators (old Sun-like stars) have weak coronae; fast rotators (young stars, T Tauri stars) have coronae 10⁴× more luminous. The qualitative mechanism — magnetic energy → corona — applies across spectral types, but the quantitative balance shifts.