A solar flare’s tiny bright cores reveal contrasting motions

A solar flare releases magnetic energy, heating the Sun’s atmosphere and driving gas into motion. But how that energy reaches the dense chromosphere, above the visible surface, remains unsettled. Measuring the size and motion of individual heated patches can help constrain explanations.

Sarah Riley and colleagues examined a minor flare on December 27, 2022, using the Daniel K. Inouye Solar Telescope, or DKIST. They found compact, spectrally enhanced cores surrounded by halos with different spectral signatures. Their paper is available as an arXiv preprint; its submission record states that it has been accepted by The Astrophysical Journal.

Sorting light into meaningful groups

The C3-class flare peaked at 20:46 UTC in active region NOAA 13176. DKIST’s Visible Spectropolarimeter recorded more than a million spectra during a roughly 93-minute scan. The researchers studied the calcium line at 8542 angstroms, whose shape provides information about chromospheric conditions and motions.

A spectrum measures brightness across wavelengths. Heating and gas motion can change a spectral line’s strength, shape and wavelength position. But interpreting those changes requires context: similar-looking profiles can arise in different background atmospheres.

The team therefore first separated spectra using brightness near the calcium line and in the nearby continuum. This helped distinguish dark pore regions from brighter plage—two different solar environments—before comparing detailed line shapes.

Within those categories, they used a hierarchical version of k-means clustering, which groups similar measurements around representative averages. Dividing each spectrum by its own continuum brightness emphasized shape rather than overall brightness. Coarser groups captured dominant features; finer subdivisions retained subtler differences. The resulting classification placed 17,422 selected flare spectra into 19 clusters.

These are categories of spectral appearance, not evidence for 19 separate physical heating mechanisms.

Cores and halos with contrasting shifts

When mapped onto the observed region, the clusters formed coherent structures rather than a random scatter. Spectra with stronger continuum-normalized line emission occupied central “kernels,” surrounded by progressively weaker groups forming halos.

The researchers identified three prominent kernels in plage and one in a pore. The pore kernel measured about 368 kilometres across along the instrument’s slit. Its extent in the scanning direction was unresolved, so this is not a complete measurement of its shape or area. Nor does it establish the smallest possible scale of flare heating.

The motion signatures differed between environments. The plage kernels had redshifted line cores corresponding to about 1.5 kilometres per second, while their halos were weakly blueshifted. The pore kernel showed a stronger blueshift, about 4.1 kilometres per second, with progressively weaker blueshifts through its inner halo. These shifts are measured relative to a quiet-region reference spectrum.

The pore surroundings also contained more varied spectral structure: 14 clusters across roughly six square arcseconds, compared with four across about 15 square arcseconds in plage.

Interpreting wavelength shifts as gas velocities requires care because these calcium profiles can contain emission peaks separated by an absorption dip. The team checked the diagnostic using 79 existing flare simulations, yielding almost 40,000 snapshots. The theoretical line-core shift tracked the average gas velocity where that core formed. However, the observational extraction method failed for some shallow or absent reversals; four observed clusters were excluded from the validated velocity interpretation.

A detailed scan is not a movie

The central limitation is time. DKIST scanned across the region rather than repeatedly measuring each location. Adjacent positions could therefore represent different stages of heating, not simply different places at one instant.

To recover context, the researchers used ultraviolet images from the Atmosphere Imaging Assembly. They compared each spectrum’s observation time with a relevant local ultraviolet brightness peak. The pore kernel was generally sampled during the rise toward that peak, whereas most plage-kernel spectra came during early decay.

That comparison cannot fully separate environment from evolution. The ultraviolet images have lower spatial resolution, mix emission from several atmospheric contributions and were smoothed over 100 seconds. Their timing estimates are meaningful on roughly one-to-two-minute scales, not as precise clocks for every calcium feature.

Different background atmospheres, including stronger magnetic fields at the pore, may help explain the contrasts. Different heating stages may also contribute. This single minor flare, observed during DKIST’s early operations, cannot decide between those possibilities or establish a universal pattern.

The next requirement is repeated, high-resolution spectroscopy throughout a flare, ideally across several spectral lines. That would let researchers follow the same small structures as they change, rather than infer their histories from a slow scan.