SolarAurora: A Cross-Domain Aurora Phenomenon

• SolarAurora is a multi-domain concept describing how solar wind energy drives auroral displays in terrestrial, solar, and stellar environments.
• It leverages methods from machine-learning forecasting and historical reconstruction to model auroral dynamics using metrics like Dst, Kp, and ROC-AUC.
• The framework unifies diverse observations and electrodynamic principles, providing actionable insights into energy transfer and predictive space weather analysis.

SolarAurora denotes a linked set of ideas centered on auroral physics as a solar, magnetospheric, and astrophysical phenomenon. In current usage it can mean the causal chain from solar activity to terrestrial auroral and ionospheric response, a causality-informed spatiotemporal forecasting framework for extreme space weather, and solar or stellar auroral analogues such as coherent radio emission above a sunspot and magnetospherically driven aurorae at the end of the stellar main sequence (Dai et al., 27 Jul 2025, Yu et al., 2023, Hallinan et al., 2015). Taken together, these usages suggest a cross-domain concept in which magnetic and plasma energy is transferred into atmospheric emission, visibility, and measurable geospace response.

1. Conceptual range

SolarAurora is not restricted to a single observational domain. In Earth studies it refers to the solar-wind–magnetosphere–ionosphere chain that produces aurora borealis and related space-weather signatures; in machine-learning studies it appears as the name of a forecasting architecture; and in solar and stellar work it labels auroral analogues outside the standard terrestrial setting.

Domain	Representative focus	Representative paper
Solar–terrestrial aurora	Long-term auroral variability, low-latitude events, extreme storms	(Vázquez et al., 2013, Hayakawa et al., 2021)
Solar coronal analogue	Long-lasting aurora-like radio emission above a sunspot	(Yu et al., 2023)
Ultracool dwarf analogue	Radio and optical aurorae on LSR J1835+3259	(Hallinan et al., 2015)
Probabilistic visibility forecasting	Two-stage occurrence/visibility decomposition	(Ge et al., 21 May 2026)
Causal spatiotemporal forecasting	Information-theoretic SolarAurora framework	(Dai et al., 27 Jul 2025)

A unifying feature across these usages is that aurora is treated as a diagnostic of large-scale electrodynamics rather than only as an optical display. This includes historical reconstruction from chronicles and geomagnetic coordinates, site-specific probability forecasts for human observation, soft- and hard-X-ray auroral morphology, coherent radio maser emission, and magnetospherically driven optical and radio emission on star-like objects.

2. Electrodynamic basis

In the terrestrial case, aurora borealis is a visible manifestation of energy transfer from the solar wind into the Earth’s magnetosphere–ionosphere system. The essential drivers are solar-wind plasma flow, embedded interplanetary magnetic field, and the geometry of coupling: when the IMF, especially its north–south component $B_z$ in GSM coordinates, turns southward, dayside reconnection opens field lines, energy is stored in the magnetotail, and substorms on 1–6 hour timescales drive particle precipitation into the auroral oval around geomagnetic latitudes $\sim 60$–$75^\circ$, especially on the nightside (Ge et al., 21 May 2026).

The electrodynamic link is carried by magnetospheric current systems, especially field-aligned currents between magnetosphere and ionosphere. Precipitating electrons collide with the upper atmosphere, excite and ionize neutrals, and generate multiwavelength emission. In planetary settings this includes highly polarized coherent radio emission, infrared, optical, ultraviolet, and X-ray components; in gas giants and ultracool dwarfs alike, auroral radio emission is associated with the electron cyclotron maser instability, with characteristic frequency tied to magnetic field through $\nu_c \approx 2.8 \times 10^6 B\,[\mathrm{Hz}]$ for $B$ in gauss (Hallinan et al., 2015, Yu et al., 2023).

A further extension arises when the solar wind becomes sub-Alfvénic. During the April 2023 storm, a Multiscale Atmosphere Geospace Environment simulation validated by AMPERE showed that when $M_A < 1$, Alfvén wings developed and northern-hemisphere planetward flowing electrons were predominantly at magnetic local times 8–13. In that state, field-aligned currents were driven by flow vorticity at the boundary of the Alfvén wings and unshocked solar wind, producing a distinct auroral precipitation geometry (Burkholder et al., 22 Feb 2025).

3. Historical reconstruction and low-latitude limits

Auroral records have long been used as a proxy for solar activity. A global compilation for 1700–1905 assembled about 27,000 auroral events with more than 80,000 observations, assigned geographic and geomagnetic coordinates to individual sites, and applied an 11-year smoothing window to isolate long-term variability. Variations of cumulative auroral counts with latitude were then used to discriminate between aurorae associated with strong or medium coronal mass ejections and those associated with high-speed streams from coronal holes, with the strongest anti-correlation with sunspots appearing above about $61$–$64^\circ$ geomagnetic latitude (Vázquez et al., 2013).

Twentieth-century archival material gives the same chain at higher observational fidelity. Japanese records around the International Geophysical Year document three extreme storms during Solar Cycle 19: March 1957 with minimum $Dst = -255$ nT, September 1957 with $Dst = -427$ nT, and February 1958 with $\sim 60$0 nT. These events yielded reconstructed equatorward auroral boundaries down to $\sim 60$1, $\sim 60$2, and $\sim 60$3 in invariant latitude, and the displays were generally reddish with occasional yellowish rays, consistent with dominant oxygen red emission mixed with green emission (Hayakawa et al., 2021).

Historical low-latitude reports define the upper tail of SolarAurora in a more dramatic way. During the 1859 Carrington event, the auroral oval expanded sufficiently that reports came from Panama and Montería, Colombia, where geomagnetic latitude was about $\sim 60$4; during the 4 February 1872 storm, aurora was observed down to about $\sim 60$5 magnetic latitude, a reconstruction gave $\sim 60$6 nT, and Secchi’s analysis linked the event to severe disturbances on telegraph systems and the transatlantic cable (Cárdenas et al., 2015, Berrilli et al., 2021).

Low latitude, however, is not synonymous with extreme storm forcing. The 15 February 1875 Rio de Janeiro event, observed at approximately $\sim 60$7, occurred with maximum daily $\sim 60$8 nT and is interpreted as a possible sporadic aurora. East Asian philological work reaches a similar conclusion from the opposite direction: terms such as “unusual rainbow” and “white rainbow” can be auroral candidates in some contexts, but entries that clearly “pierce” the Sun or Moon remain more naturally explained as halo phenomena (Oliveira et al., 2020, Hayakawa et al., 2016).

4. Solar, planetary, and stellar analogues

A solar-coronal SolarAurora analogue was identified in long-lasting radio emission above a sunspot observed on 2016-04-09. The source was seen from 1–1.7 GHz with the VLA and down to 245 MHz with whole-Sun radio monitors, showed nearly 100% right-hand circular polarization between 1.0–1.4 GHz, and had inferred brightness temperatures of about $\sim 60$9 K at 1 GHz and $75^\circ$0 K at 245 MHz for a compact source. Imaging, magnetic extrapolation, and propagation arguments led to an interpretation in terms of second-harmonic O-mode electron cyclotron maser emission above a strong, converging sunspot field, powered by energetic electrons supplied by recurring nearby flares (Yu et al., 2023).

At the end of the stellar main sequence, the M8.5 dwarf LSR J1835+3259 established a stellar-scale auroral analogue. Simultaneous radio and optical spectroscopy showed periodic coherent radio bursts together with Balmer-line and continuum variability on a 2.84 hour rotation period. ECMI frequencies implied magnetic fields of about $75^\circ$1–$75^\circ$2 kG, the radio auroral power was estimated at $75^\circ$3 W, the precipitating electron beam power at $75^\circ$4–$75^\circ$5 W, and the dissipated power was at least four orders of magnitude larger than what is produced in the Jovian magnetosphere. The key physical distinction from solar-type activity is that the energy source is interpreted as large-scale magnetospheric electrodynamics rather than lower-atmosphere reconnection alone (Hallinan et al., 2015).

Mars provides a planetary counterexample in which auroral processes operate without an intrinsic global magnetic field. Because Mars instead has distributed crustal fields, its auroras have been classified as diffuse, discrete, and proton aurora, and Hope observations added discrete sinuous aurora and patchy proton aurora. This suggests that SolarAurora-like behavior can persist across dipolar magnetospheres, induced magnetospheres, the solar corona, and star-like objects, while the topology of the field and the source population of precipitating particles vary strongly from case to case (Atri et al., 2022).

5. Forecasting, visibility, and causal modeling

Operationally, auroral occurrence and auroral visibility are different targets. Aurora Hunter formalizes this by factorizing

$75^\circ$6

Its Stage 1 uses XGBoost with 51 physics-driven features trained on joint Tromso+Kiruna data of about 16,600 hourly samples from 2015–2023, while Stage 2 uses logistic regression with 21 cloud-cover and lunar-illumination features trained only on aurora-occurring hours. The cascade reaches ROC-AUC 0.937 on the Tromso test set and 0.905 on independent Kiruna 2024 data, improving a single-stage baseline by +0.087; SHAP analysis identifies the $75^\circ$7 nightside interaction, MLT position, and auroral oval distance as the dominant predictors, contributing 39% combined (Ge et al., 21 May 2026).

As a proper noun, SolarAurora is also a causality-informed spatiotemporal machine-learning framework for extreme space weather. It combines partial information decomposition, synergy flux, and a Spherical Fourier Neural Operator backbone with causal consistency constraints, and was trained on 1980–2024 datasets spanning solar proxies, solar wind, IMF, geomagnetic indices, and global VTEC. The framework identifies a hierarchical chain from solar proxies through solar wind and IMF to $75^\circ$8, $75^\circ$9, and ionospheric response; stable delays include $\nu_c \approx 2.8 \times 10^6 B\,[\mathrm{Hz}]$0 of about 4.5 hours and magnetosphere-to-ionosphere links of about 4–6 hours. For May 2024, reported 3-hour-ahead MAPE reductions relative to Swin-T are 9.02% for F10.7, 22.43% for $\nu_c \approx 2.8 \times 10^6 B\,[\mathrm{Hz}]$1, and 4.17% for $\nu_c \approx 2.8 \times 10^6 B\,[\mathrm{Hz}]$2, while auroral spatial causality patterns intensify in the northern auroral region and coincide with high ROTI, linking causality maps to Arctic space-weather threats (Dai et al., 27 Jul 2025).

The two frameworks therefore solve different layers of the same problem. Aurora Hunter asks whether an observer at a specified site can actually see aurora through clouds and moonlight, whereas SolarAurora asks how information and influence propagate through the solar–magnetosphere–ionosphere system and how that structure can improve global forecasting. This suggests a layered SolarAurora methodology in which physical occurrence, spatial impact, and local visibility are modeled separately when the application requires it.

6. Asymmetry, instrumentation, and unresolved problems

AuroraMag extends SolarAurora from theory and inference to measurement architecture. The mission concept uses two small satellites carrying identical auroral X-ray imagers, an in situ particle detector, a magnetometer pair, and an electron temperature analyzer, with the explicit goal of observing Earth’s northern and southern auroral ovals simultaneously while measuring particles, fields, and temperature in situ. Its primary scientific targets are the formation, morphology, and hemispherical asymmetries of X-ray aurora, precipitating particle fluxes, Solar Energetic Particles, currents, and cusp dynamics, and it is presented as the first dedicated twin-spacecraft mission of such kind for simultaneous study of hemispheric asymmetries in solar-wind–magnetosphere coupling (Bhaskar et al., 2024).

Several major questions remain open. In ultracool dwarfs, the electrodynamic engine responsible for the auroral current system may be corotation breakdown, an orbiting satellite–magnetosphere interaction, or reconnection with the ambient medium, and present observations do not uniquely distinguish among them (Hallinan et al., 2015). In the solar-coronal analogue above a sunspot, the exact electron distribution, the role of parallel electric fields and Alfvén turbulence, and the occurrence rate of such events remain uncertain (Yu et al., 2023). Even in the terrestrial case, interpretation can be geometry-limited: the April 2023 Hanle event was seen from $\nu_c \approx 2.8 \times 10^6 B\,[\mathrm{Hz}]$3N although the equatorward boundary of the aurora remained beyond $\nu_c \approx 2.8 \times 10^6 B\,[\mathrm{Hz}]$4N geographic latitude, because red emissions at about 700–950 km altitude over roughly $\nu_c \approx 2.8 \times 10^6 B\,[\mathrm{Hz}]$5N could still be visible from Hanle at elevations of about $\nu_c \approx 2.8 \times 10^6 B\,[\mathrm{Hz}]$6–$\nu_c \approx 2.8 \times 10^6 B\,[\mathrm{Hz}]$7 (Vichare et al., 2024).

In this broader sense, SolarAurora is not a single phenomenon but a research program connecting solar forcing, magnetospheric electrodynamics, atmospheric emission, visibility, and forecasting across Earth, the Sun, planets, and star-like objects. A plausible implication is that the same core plasma processes—reconnection, field-aligned currents, particle precipitation, wave–particle interaction, and coherent cyclotron emission—recur across very different magnetic geometries, while the observable manifestation depends on field topology, plasma source, atmospheric composition, and line-of-sight geometry.