The dunes were first described in the scientific literature in 2020, when a study led by Maria Palmroth used two nearly simultaneous photographs taken by amateur observers to establish that the auroral emission associated with the dunes was produced at an altitude of approximately 100 kilometers. That altitude places the phenomenon in the lower thermosphere, a notoriously difficult region to measure directly. Follow-up work in 2021 revealed the remarkable scale of a single dune event: the wave-like stripes spanned more than 1,500 kilometers of horizontal distance, from western Finland to Scotland, and remained visible for at least four hours. Those studies also found that the dunes were associated with electron precipitation in the 10 to 20 kiloelectronvolt range and with a large temperature inversion layer just below the mesopause, the boundary region roughly 85 to 100 kilometers up where temperatures reach their coldest point in the entire atmosphere.

The leading hypothesis is that the dunes are an optical signature of an atmospheric wave propagating horizontally through the diffuse aurora near the mesopause. As the wave modulates the density of atomic oxygen, the ingredient responsible for the green auroral emission, it imprints its crests and troughs onto the glow, producing the characteristic stripes. The prime candidate is the mesospheric bore, a sharp wavefront followed by a train of trailing crests that can only propagate within a ducting structure, such as a temperature inversion layer or a wind shear. Bores have horizontal wavelengths of 20 to over 100 kilometers, compatible with the 30 to 45 kilometer pseudo-wavelength measured in the few dune events studied so far. But alternatives exist: some researchers have proposed acoustic waves ducted through dusty plasma produced by meteor showers, while others have suggested a purely magnetospheric origin, in which waves near the plasmapause spatially modulate the flux of precipitating particles. Without statistics, the debate could not be settled.

The new survey tackles that gap by mining Skywarden, an observation database created and maintained by the Ursa Astronomical Association in Finland. The researchers searched the archive from January 2000 to December 2025, restricting the search to reports containing at least one photograph showing the dunes. Each candidate observation then passed through a three-step verification process: the observers themselves flagged possible dune structures, the Skywarden moderating team reviewed the auroral forms present in each submission, and finally two of the study’s authors visually inspected every report, discarding cases where clouds or airglow structures could mimic the dunes. The result was a curated set of 308 dune observations, made during 61 distinct events by 183 named citizen scientists from Northern Europe, North America, Australia, and New Zealand. As a baseline for comparison, the team also retrieved nearly 10,000 timestamped aurora reports from the same database.

The statistical patterns that emerged are striking. Ninety-two percent of the dune reports came from observers below 61 degrees geomagnetic latitude, compared with 74 percent for all aurora observations, suggesting that the dunes frequently occur in the equatorward, subauroral part of the auroral oval. The local time distribution is even more distinctive: 89 percent of dune observations were made before magnetic midnight, with a sharp peak between 21:00 and 22:00 magnetic local time, roughly two hours earlier than the peak for aurora in general. Permutation tests with false discovery rate correction confirmed that both the local time difference and the monthly distribution difference are statistically significant, while the solar wind and interplanetary magnetic field parameters showed no significant differences between dune events and typical aurora. In other words, the dunes do not appear to require a special kind of solar wind driving.

What the dunes do require, apparently, is geomagnetic activity. The SuperMAG SME index, a proxy for auroral electrojet intensity derived from ground magnetometers, had a median value of 871 nanotesla during dune observations, significantly higher than the median of 507 nanotesla for all aurora reports. The SMR ring current index was likewise skewed toward more disturbed values, with a median of minus 49 nanotesla, though the dunes do not necessarily demand a full-blown geomagnetic storm. Individual events could span more than 5 degrees of geomagnetic latitude, and combining multiple observers’ reports showed that about a third of dune events lasted more than two hours, with the longest, observed from New Zealand and Australia during the great geomagnetic storm of 11 May 2024, persisting for 6.9 hours. Seasonally, the events peaked in October, with a secondary peak in March, a pattern shaped by darkness, cloudiness over Finland, and the equinoctial enhancement of geomagnetic activity known as the Russell-McPherron effect.

Perhaps the most technically revealing part of the study concerns ionospheric currents. For 15 subsets of dune observations made by multiple photographers in Fennoscandia, the team applied the Spherical Elementary Current System method to data from the IMAGE magnetometer network, reconstructing the equivalent ionospheric currents with a spatial resolution of about 100 kilometers. The analysis showed that the dunes were consistently associated with strong auroral electrojet signatures: in 10 of the 15 cases they occurred within the eastward electrojet, matching the original 2020 discovery, but in one case they appeared within the westward electrojet, and in four cases they persisted through the transition from eastward to westward current. Crucially, every analyzed event took place in the vicinity of the Harang discontinuity, the boundary region on the nightside where the large-scale ionospheric current pattern reverses direction.

These current patterns, combined with the strong pre-midnight preference, point toward an intriguing possibility: that precipitating protons may play a key role in creating the diffuse green background against which the dunes appear. Proton precipitation produces secondary electrons that excite atmospheric constituents, generating diffuse emissions that concentrate on the equatorward side of the auroral oval in the dusk sector, precisely where the dunes are most often reported. The authors caution that confirming this requires detailed case studies with satellite particle measurements or ground-based spectrometers. Meanwhile, the seasonal anomalies, particularly the surplus of October events and the scarcity of September and February events relative to general aurora activity, hint that middle-atmospheric dynamics, including gravity waves, mesospheric bores, and the seasonal behavior of the stratospheric polar vortex, may also shape when the dunes can form. Short-lived events may trace gravity waves, while long-lived displays may require the sustained duct of a mesospheric bore.

Beyond the physics, the study stands as a milestone for citizen science in space physics. The dunes are low-contrast structures in a dim auroral form, best seen at low elevations with the narrow field of view of a commercial camera, conditions under which all-sky imaging networks struggle. The researchers acknowledge biases in their dataset, from the dominance of Finnish observers to uncertainties of up to 463 kilometers in the true location of the dunes relative to the photographers, but they argue that the growing quality of citizen science data, supported by field guides, moderator review, and coordinated international networks, is transforming amateur photographs into a legitimate scientific record. As aurora chasing communities around the world continue to grow, the sky itself is becoming a distributed observatory, and phenomena once invisible to science are stepping into the light.

Subject of Research: Statistical survey of the dune aurora phenomenon using citizen science observations

Article Title: Dune aurora: survey from a citizen science database

Article References: Grandin, M., Juusola, L., Partamies, N., Bruus, E., Rautiainen, J., Lach, D., Jia, J., van de Kamp, M., Karvinen, E., Kauristie, K., & Hoppe, T. (2026). Dune aurora: survey from a citizen science database. Annales Geophysicae, 44(2), 855-880. https://doi.org/10.5194/angeo-44-855-2026

Image Credits: AI Generated

DOI: 10.5194/angeo-44-855-2026

Keywords: dune aurora, citizen science, aurora, mesospheric bore, ionosphere, thermosphere, geomagnetic storm, auroral electrojet, Harang discontinuity, Skywarden, space physics, atmospheric waves