The study, led by Etienne Gavazzi of UiT The Arctic University of Norway in Tromsø and published in Annales Geophysicae, mines measurements from the VISIONS-2 sounding rocket mission. Two rockets were launched from Ny-Ålesund on Svalbard on 7 December 2018, flying south-west through the dayside auroral region at around 13.5 magnetic local time. The lower of the two payloads, which reached an apogee of just over 600 kilometers, carried a top-hat electrostatic analyser capable of sweeping through 49 energy steps from 3 electronvolts to 28.6 kiloelectronvolts every 50 milliseconds, while simultaneously resolving the directions of incoming electrons across twenty pitch-angle channels. That combination of speed and angular resolution proved decisive.

The physical culprit behind these auroras is a family of electromagnetic disturbances known as dispersive Alfvén waves. Ordinary magnetohydrodynamic Alfvén waves cannot carry a parallel electric field, the ingredient needed to accelerate electrons along magnetic field lines. But when the waves acquire small perpendicular scales, comparable to the electron inertial length in the tenuous plasma below a few Earth radii, or to the ion acoustic gyroradius farther out, they become dispersive and develop exactly that field-aligned electric field. Through resonant processes, electrons surfing these waves can be driven to parallel velocities approaching, and in some regimes exceeding, the local Alfvén speed, producing the kiloelectronvolt electrons that light up the sky. Previous spacecraft statistics suggested that Alfvén waves power on average a quarter to nearly forty percent of all auroral electron energy, dominating under active conditions near noon and midnight.

The fingerprint of this wave-driven precipitation is distinctive. In time-energy spectrograms, the electrons arrive broadband in energy and dispersed in time: the fastest, most energetic electrons reach the spacecraft first, with progressively slower ones trailing behind. The VISIONS-2 low-flyer recorded dozens of such dispersed structures during its flight, with peak energies from a few hundred electronvolts to several kiloelectronvolts and durations typically under half a second. Crucially, the same intervals showed semicircular dispersions in pitch-angle, where field-aligned electrons arrived first and arrival time grew with angle, together with broadband electromagnetic fluctuations and bursts of downward Poynting flux. A cross-spectral analysis of the perpendicular electric and magnetic fields yielded a ratio of about 2500 to 3000 kilometers per second between 1 and 8 hertz, closely matching the local Alfvén speed of roughly 2730 kilometers per second, and cross-phases consistent with waves propagating downward into the ionosphere.

To convert these fleeting signatures into altitudes, the team applied the classical time-of-flight technique: assuming electrons of all energies were released simultaneously from a single height, a linear fit of arrival time against inverse velocity yields the source distance. They also tested a quadratic fit, previously shown to describe dayside dispersions better, and introduced a logarithmic fit that appears as a straight line on semi-logarithmic energy-time plots. Using the corrected Akaike Information Criterion to compare models across 29 well-resolved events, the non-linear fits won out overwhelmingly, with 14 events best described by the logarithmic curve, 12 by the quadratic, and only 3 by the linear model. That statistical preference itself carries meaning: it supports a picture in which acceleration is smeared across a range of altitudes as the wave travels downward, rather than confined to a single point.

The real methodological innovation, however, lies in a second approach that the authors developed to escape the simultaneous-release assumption altogether. Instead of fitting energy-time curves, they built a forward model of pitch-angle-time dispersions. Electrons of a single energy are launched from a candidate source altitude with a spread of pitch angles, their trajectories computed under magnetic mirroring using the IGRF-14 field model, and their arrival times convolved with a Gaussian injection profile and the detector’s sampling behaviour. Three parameters, source altitude, injection duration, and a time offset, are then adjusted for each energy channel independently by minimizing the squared error between modelled and observed dispersions. Because each energy bin receives its own release height, the method relaxes the core assumption of the energy-time techniques while exploiting the rocket’s unusually fine angular resolution.

The two families of methods, resting on different assumptions, converged. For most analysed structures, the inferred source altitudes fell between 1000 and 3000 kilometers, with a clear trend across events that higher-energy electrons were released higher up: a few hundred electronvolts mapped to roughly 1000 to 3000 kilometers, while kiloelectronvolt electrons traced back to about 2000 to 5000 kilometers. Where the methods disagreed, the differences rarely exceeded 1000 kilometers. The fitted time offsets also increased systematically for lower-energy electrons, indicating later release times, exactly what one expects if the acceleration region rides downward with a propagating wave. One event at 500 seconds flight time even yielded a source altitude near 6400 kilometers for 4300-electronvolt electrons, with modelled source pitch angles confined within about 25 degrees of field alignment.

Placed against theory, the results sit at the lower end of expectations but remain physically coherent. When the team compared their altitude-energy scatter with inertial Alfvén wave velocity profiles computed from published plasma density models, the observations aligned best with profiles featuring relatively small oxygen ion scale heights, meaning the plasma transitions from heavy O+ dominance to light H+ dominance at comparatively low altitudes. Almost none of the events matched the outside-cavity profile of Lysak and Song, while many were consistent with profiles associated with density cavities, the depleted plasma regions long linked to intense parallel electric fields in auroral zone observations. The event-to-event variability in source altitude, even within a single rocket flight, points to genuinely different plasma and wave conditions along neighbouring magnetic field lines, possibly structured by the ionospheric Alfvén resonator or by the aftermath of earlier precipitation.

Perhaps the most exciting prospect is what the technique makes measurable. By fitting the inferred release altitudes and energies to a parametrized inertial Alfvén speed profile, the authors showed that the oxygen and hydrogen density profiles and the perpendicular wavelength of the wave can be reconstructed from rocket data alone. Applied across their events, about two-thirds produced reasonably constrained fits, with best-fit parameters remarkably close to values reported in earlier statistical studies. Because the topside ionosphere near the O+-to-H+ transition is one of the most poorly sampled regions of near-Earth space, a method that turns a brief rocket traversal into a density profile and a wave scale is a genuinely valuable addition to the toolkit. Future instruments with higher time, energy, and angular resolution could sharpen these estimates further, and the approach could be extended to satellites, trading some precision for coverage across a far wider range of geomagnetic conditions.

For now, the study delivers something auroral physics has long wanted: a consistent, multi-method measurement of where the sky’s most restless lights get their power. The electrons that set the dayside aurora ablaze appear to be flung Earthward from effective release altitudes of one to three thousand kilometers most of the time, energized progressively by inertial Alfvén waves as those waves sweep down magnetic field lines. Each dispersed burst recorded over half a second of rocket flight is thus a snapshot of a wave-particle interaction unfolding thousands of kilometers overhead, and with the new pitch-angle technique, that snapshot can now be read as a probe of the plasma environment itself, turning the aurora from a spectacle into an instrument.

Subject of Research: Estimating the source altitudes of Alfvénic auroral electron acceleration in the dayside ionosphere using sounding rocket dispersion measurements

Article Title: Estimating the source altitude of auroral precipitation from dispersed Alfvén waves in the dayside ionosphere

Article References: Gavazzi, E., Spicher, A., Gustavsson, B., Vierinen, J., Clemmons, J., Pfaff, R., & Rowland, D. (2026). Estimating the source altitude of auroral precipitation from dispersed Alfvén waves in the dayside ionosphere. Annales Geophysicae, 44(2), 903-920. https://doi.org/10.5194/angeo-44-903-2026

Image Credits: AI Generated

DOI: 10.5194/angeo-44-903-2026

Keywords: aurora, Alfvén waves, sounding rockets, VISIONS-2, space plasma physics, electron acceleration, dayside ionosphere, pitch-angle dispersion, time-of-flight analysis, magnetosphere-ionosphere coupling, plasma density profiles, Annales Geophysicae

The research, published in Astrophysics and Space Science, drew on high-resolution dynamic spectra from the e-CALLISTO network, a worldwide chain of inexpensive radio spectrometers that monitors the Sun continuously from multiple ground-based stations. Between June 2020 and December 2024, the team identified 18 J-type and 20 U-type bursts in the decametric frequency range of 10 to 200 megahertz, where these burst types are most prominent. Each event was screened through a multi-step process: candidate bursts were spotted visually in dynamic spectrograms, validated for consistent morphology and timing across independent stations, and rejected if their frequency drift did not follow the expected trajectory — a low-to-high frequency path for J-types, or a full low-to-high-to-low reversal for U-types.

The distinction between the two burst types lies in the geometry of the magnetic structures that guide the electron beams. A J-type burst appears in a dynamic spectrum as a hook-like feature: the emission frequency drifts downward as electrons climb through the corona, then terminates abruptly without reversing, suggesting that the beam dissipates or scatters somewhere along a loop anchored at a single footpoint. A U-type burst, by contrast, shows a symmetric frequency evolution — the drift reverses cleanly at a minimum frequency, marking the moment the electron beam passes over the apex of a closed loop and begins descending the opposite leg toward the solar surface. Because the emission frequency at any instant reflects the local plasma density through the plasma frequency relation, the shape of the burst on a spectrogram is effectively a map of the density — and therefore the height — along the electron’s path.

To convert those spectral curves into physical quantities, the researchers traced the backbone of each burst and applied the Newkirk density model, a classical empirical description of coronal electron density that falls off with heliocentric distance. From the frequency drift rate, they derived the radial velocity of the electron beams and the heights at which the emission originated. The results revealed a striking dichotomy. J-type bursts reached apex heights between 1.2 and 2.5 solar radii, with a mean near 1.6 solar radii, and displayed steep negative drift rates in their descending legs, typically ranging up to 14.6 megahertz per second. U-type bursts extended dramatically higher — up to 3.9 solar radii, with a mean around 2.5 solar radii — and showed smoother, more gradual drift patterns on their ascending branches, with drift rates spanning a broad range from 0.5 to 139.6 megahertz per second.

Those asymmetries carry real physical meaning. The differences between the ascending and descending branches of U-type bursts suggest that electron propagation is not symmetric around the loop apex, reflecting unequal density gradients, magnetic geometries, or beam conditions along the two legs of the closed structure. For J-type bursts, the steeper drifts point to rapid electron deceleration and confinement within short, low-lying loops, consistent with sharp density gradients and localized acceleration near compact reconnection sites. In effect, the spectral morphology of each burst encodes the geometry of the loop that produced it: J-types trace compact, often asymmetric structures, while U-types map long-lived, large-scale closed loops.

Perhaps the most surprising finding concerns the role of solar flares. Cross-referencing all 38 events with the GOES soft X-ray flare catalog, the team found that only 26 were associated with flares at all — and roughly 30 percent of the bursts occurred without any reported flare. Among the flare-associated events, C-class flares dominated, accounting for about 56 percent of J-type and 45 percent of U-type bursts, while a handful coincided with M-class flares and just one U-type event matched a B4.3 flare. No X-class flares appeared in the sample. This means the electron acceleration responsible for these bursts is not exclusively tied to impulsive flare energy release; smaller-scale reconnection, active region restructuring, or turbulence within coronal loops can also launch electron beams energetic enough to produce detectable radio emission.

To validate the loop heights inferred from the density model, the team turned to an entirely independent method: potential field source surface extrapolations built from photospheric magnetograms. This technique assumes the coronal field is current-free and divergence-free between the solar surface and a spherical source surface fixed at 2.5 solar radii, beyond which field lines are forced to be radial. Using GONG synoptic magnetograms matched to each event’s Carrington rotation, and the open-source pfsspy package wrapping a finite-difference solver, the researchers traced closed field lines through the three-dimensional solution and calculated the geometric height of each loop’s arc. Extreme ultraviolet images from SDO’s Atmospheric Imaging Assembly provided additional context, identifying the active regions and loop footpoints associated with the radio sources.

The comparison between the two methods was revealing but constrained. Of the 38 events, 15 had apex heights above the 2.5-solar-radius source surface boundary and lay outside the PFSS domain entirely, while four more originated near the solar limb where the magnetogram input is poorly constrained. That left 19 events with valid estimates from both approaches. For this subset, PFSS yielded loop heights from 0.36 to 2.5 solar radii with a mean near 1.6, while the Newkirk inversion ranged from 1.2 to 4.3 solar radii with a mean near 2.5 — a systematic offset of roughly 0.9 solar radii. Yet the event-to-event differences stayed within about half a solar radius, indicating reasonable agreement despite the models’ fundamentally different assumptions. The pattern was clear: PFSS reliably captured short, low-altitude loops typical of J-type bursts, while systematically underestimating the extended structures associated with U-type bursts, which the density-based method recovered up to 4.3 solar radii.

The study also extends the known scale of these phenomena. Earlier imaging spectroscopy with LOFAR had placed U- and J-type loop altitudes at roughly one solar radius; the new statistical analysis of 38 events pushes that ceiling considerably higher. Several U-type bursts even displayed double-lane or complex frequency structures, hinting at multiple electron injections or partial reflections within loops — features that make U-bursts uniquely valuable for investigating multi-episode acceleration and electron trapping, processes largely inaccessible through type III studies alone.

For space weather research, the implications are tangible. Closed-field electron acceleration governs how energetic particles are stored, scattered, or released from the corona, and knowing whether a burst originates from a compact low-lying loop or a sprawling structure extending several solar radii outward helps assess the connectivity between the Sun and interplanetary space. The authors argue that combining density-model inversions with magnetic extrapolations offers a robust dual diagnostic, and they point toward the next generation of instruments — LOFAR, the Murchison Widefield Array, the upcoming Square Kilometre Array, and in-situ measurements from Parker Solar Probe and Solar Orbiter — as the tools that will sharpen this picture further. As Solar Cycle 25 continues toward its maximum, every hooked and U-turned streak on a radio spectrogram becomes a free measurement of the corona’s hidden architecture, delivered by electrons that race along magnetic rails and broadcast their journey across the solar system.

Subject of Research: Closed-field electron acceleration and magnetic loop diagnostics using J- and U-type solar radio bursts in the solar corona

Article Title: Closed-field electron acceleration in the solar corona: insights from J- and U-type solar radio bursts

Article References: Soni, S. L., Luthra, V., Singh, J., Suklan, R., Bhatt, T. K. N., & Jain, R. (2026). Closed-field electron acceleration in the solar corona: insights from J- and U-type solar radio bursts. Astrophysics and Space Science, 371(9), Article 111. https://doi.org/10.1007/s10509-026-04639-4

Image Credits: AI Generated

DOI: 10.1007/s10509-026-04639-4

Keywords: solar radio bursts, J-type bursts, U-type bursts, solar corona, magnetic loops, electron acceleration, Solar Cycle 25, e-CALLISTO, PFSS extrapolation, Newkirk density model, space weather, solar flares