The instrument in question, JUICE-UVS, is an imaging spectrograph sensitive to extreme- and far-ultraviolet light between 50 and 204 nanometers, built by the Southwest Research Institute as the fifth in a family of six UV spectrographs flown on planetary missions. Light enters through one of three apertures: a main airglow port, a smaller high-resolution port tucked behind a deployable door, and a pinhole solar port offset by 60 degrees that attenuates the Sun’s glare so the instrument can eventually observe at 5 astronomical units from the Sun. An off-axis paraboloidal mirror focuses incoming light onto a slit measuring 0.1 degrees by 7.3 degrees, after which a toroidal diffraction grating spreads the UV spectrum across a microchannel plate detector coated with UV-sensitive cesium iodide. Because the optics and housing are all machined from aluminum 6061, the system is athermal, holding its focus without active temperature control.

Before the LEGA encounter, the team had verified basic performance by observing UV-bright stars and the faint glow of interplanetary hydrogen and helium. But stars are point sources, and a full characterization of the instrument’s response across all wavelengths and all positions in its field of view demands extended targets with discrete spectral features spanning the entire bandpass. That is precisely what the Moon and Earth delivered. During the lunar flyby, JUICE-UVS captured sunlight reflected from the Moon’s surface beginning roughly two minutes after the field of view crossed the terminator onto the sunlit hemisphere. Light levels rose so quickly that only the first 30-second exposure could use the main aperture; the rest of the sequence shifted to the high-resolution port to avoid overwhelming the detector.

The lunar data served two calibration purposes at once. First, the team summed counts over the 155 to 197 nanometer range, deliberately avoiding the intensely bright Lyman-alpha line at 121.6 nanometers and a reflected solar oxygen line at 130.4 nanometers where the flux temporarily saturated the detector’s ability to process every photon. The resulting map traced a narrow strip across craters including Kapteyn E, Langrenus W and Langrenus N, revealing sharp contrasts between sunlit and shadowed terrain. Because the spacecraft’s motion carried the slit mostly along its length rather than across it, the mapped region was narrow, but matching the observed crater positions to their known locations confirmed that the instrument pointing model used by the mission is accurate. The team could even resolve the shadow inside the 12.5-kilometer-wide Langrenus N crater.

Second, the lunar spectra provided a check on the instrument’s effective area, the quantity that converts raw counts into physical brightness. After correcting for exposure time, slit solid angle, viewing geometry using the Lommel-Seeliger photometric function, and dividing by a solar spectrum measured the same day by NASA’s TIMED/SEE instrument, the researchers obtained far-ultraviolet reflectance spectra for two regions: terrain near Langrenus crater and a patch of Mare Fecunditatis. Comparing these with published spectra from LAMP, the Lyman Alpha Mapping Project on NASA’s Lunar Reconnaissance Orbiter, showed excellent agreement at wavelengths below 165 nanometers. A possible absorption feature near 170 nanometers, seen in most LAMP spectra but absent from the JUICE data, remains an open question, potentially reflecting real differences in surface properties, calibration uncertainties in either instrument, or the roughly 16 percent uncertainty in the solar reference spectrum over that interval.

The Earth flyby offered an even richer calibration target. For nearly 35 minutes, JUICE-UVS observed from the nightside at an altitude of about 13,550 kilometers, across the dawn terminator, and onto the brilliant dayside, finishing with a view of the sunlit limb. Earth’s dayglow spectrum is crowded with emission lines from hydrogen, helium, oxygen and nitrogen at precisely known wavelengths, making it an ideal ruler for wavelength calibration. Using a high-resolution pixel list exposure of the dayside, the team fitted Gaussian profiles to a set of strong, unblended emission lines at multiple positions along the slit, then interpolated the results into a two-dimensional wavelength solution best described by a fifth-order polynomial. The measured pixel size varies from about 0.07 nanometers at the edges of the bandpass to about 0.11 nanometers near the center, consistent with pre-launch calibration.

The airglow measurements themselves carried scientific weight. Tracking the brightness of hydrogen Lyman-alpha and the two neutral oxygen emissions at 130.4 and 135.6 nanometers, the team found nightside values closely matching a classic 1991 compilation by R. R. Meier, while dayside peaks reached 21 kilorayleighs at 130.4 nanometers and 4.4 kilorayleighs at 135.6 nanometers. The 135.6-nanometer value exceeded contemporaneous measurements from TIMED’s Global Ultraviolet Imager, and the discrepancy is instructive rather than alarming: as the instrument’s boresight swung away from nadir, it sampled longer atmospheric columns. The optically thin 135.6-nanometer emission kept brightening with increasing off-nadir angle, while the optically thick 130.4-nanometer emission plateaued, driving the ratio of the two lines below the nadir-pointed reference value. This viewing-geometry effect, first modeled in the 1990s from Dynamics Explorer 1 data, is exactly the diagnostic JUICE-UVS will exploit at Jupiter’s moons.

Three days after the flyby, a 360-degree spacecraft roll gave UVS one final Earth observation: a 7.5-degree swath covering altitudes between roughly 2 and 15 Earth radii. The image captured the Earth’s extended hydrogen geocorona glowing in Lyman-alpha, along with the Pleiades and UV-bright stars in Taurus. The measured stellar positions matched catalog predictions almost perfectly, and the geocoronal brightness profile agreed with decades of prior observations. Searches for high-altitude helium and oxygen emissions came up empty at detectable levels. Together, these checks validated the instrument’s wavelength scale, effective area, and pointing with a completeness that stars alone could never provide.

Most importantly, the LEGA campaign was a dress rehearsal for Jupiter. During future flybys of Ganymede, Europa and Callisto, UVS will perform disk scans and limb stares to map surface ice, tenuous atmospheres, and extended hydrogen and oxygen coronas. The oxygen line ratio technique demonstrated at Earth will distinguish O2-dominated atmospheres from those containing water vapor or atomic oxygen, the same method that revealed a localized water atmosphere on Ganymede’s subsolar point. The lunar map, acquired from an altitude within the range planned for real flybys, previews the instrument’s mapping resolution of roughly 4 to 6 kilometers from 1,100 kilometers up. One technique remains undemonstrated: absorption spectroscopy via stellar and solar occultations, which the team has flagged as the highest-priority UVS activity for the final Earth gravity assist in January 2029. With the instrument confirmed healthy and its calibration sharpened, JUICE-UVS now waits, pointed toward a 2031 arrival at the largest planet in the solar system.

Subject of Research: Ultraviolet spectrograph observations and in-flight calibration of ESA's JUICE spacecraft during the 2024 lunar-Earth gravity assist

Article Title: Ultraviolet observations of the Earth and Moon during the JUICE Lunar-Earth flyby

Article References: Molyneux, P. M., Davis, M. W., Giles, R. S., Greathouse, T. K., Retherford, K. D., Versteeg, M. H., Ferrell, S. A., Becker, T. M., Gladstone, G. R., Trantham, B. J., Kammer, J. A., Freeman, M. A., & Persyn, S. C. (2026). Ultraviolet observations of the Earth and Moon during the JUICE Lunar-Earth flyby. Annales Geophysicae, 44(2), 1093-1108. https://doi.org/10.5194/angeo-44-1093-2026

Image Credits: AI Generated

DOI: 10.5194/angeo-44-1093-2026

Keywords: JUICE, JUICE-UVS, ultraviolet spectroscopy, lunar-Earth gravity assist, Moon, Earth airglow, geocorona, instrument calibration, Jupiter icy moons, ESA, Southwest Research Institute, Ganymede

The key quantity is the Lyman continuum escape fraction, written by astronomers as fesc^LyC. Lyman continuum photons are the highest-energy ultraviolet photons that massive stars can produce, with wavelengths shorter than 912 angstroms, the ionization threshold of hydrogen. Only photons at these energies can ionize intergalactic hydrogen, yet within a star-forming galaxy they face a gauntlet of obstacles: vast reservoirs of neutral hydrogen and interstellar dust that absorb them long before they can reach the galactic frontier. The abundance of ultraviolet-bright, compact star-forming galaxies observed at early cosmic times by JWST suggests that the sheer output of ionizing photons from massive stars alone could have reionized the universe, but only if a sufficiently large fraction of those photons actually escaped. Measuring that fraction directly at high redshift is nearly impossible, because intervening hydrogen absorbs the Lyman continuum before it reaches our telescopes.

This is why local analogues matter. The Low-redshift Lyman Continuum Survey, or LzLCS, and its extension known as LzLCS+ have assembled a sample of nearby compact, highly star-forming galaxies whose properties resemble those of reionization-era systems: intense star formation packed into regions smaller than a kiloparsec, low metallicity, and high ionization states of their gas. Crucially, for these nearby objects, the Hubble Space Telescope’s Cosmic Origins Spectrograph can detect escaping Lyman continuum photons directly, giving researchers ground truth about fesc^LyC that no high-redshift survey can currently provide.

In the new work, Carr and colleagues focused on six of these local galaxies that are otherwise strikingly similar to one another, but that span a range of starburst ages and Lyman continuum escape fractions. The six objects, with designations such as J103344+635317, J105331+523753, J144010+461937 and J154050+572442, were observed with medium-resolution far-ultraviolet spectroscopy, drawing on archival HST programmes and new imaging from the LaCOS survey. The power of the approach lies in its simultaneity: within a single spectrum, the team could read off both the age of the stellar population and the kinematics of the gas, allowing them to trace the launch of a galactic wind through individual absorption lines while directly measuring how many ionizing photons were escaping.

The technical trick is that different ultraviolet spectral features encode different physical information. Stellar-wind doublets such as N V at 1239 and 1243 angstroms and O VI at 1032 and 1038 angstroms are exquisitely sensitive to the ages of the youngest stellar populations, because only the most massive stars launch powerful radiation-driven winds, and those stars die within a few million years. Photospheric features, by contrast, trace stellar metallicity. On the galactic side, absorption lines imprinted by outflowing interstellar gas, including H I Ly-beta at 1026 angstroms and C II at 1036 angstroms, record the velocity structure and column density of material along the line of sight. By fitting stellar population synthesis models with the FiCUS spectral fitting code and modeling radiative transfer with the publicly available SALT semi-analytic line transfer code, the team disentangled these stellar and interstellar components.

The result is a clear evolutionary sequence. In the youngest systems, with stellar populations younger than roughly three million years, Lyman continuum escape peaks. At this stage, the galaxy is flooded with intense radiation from its most massive stars, and strong stellar winds from those stars are carving channels through the neutral gas, but a fast large-scale galactic wind has not yet developed. The youngest starbursts are, somewhat counterintuitively, the best leakers. Then, as the starburst ages past three million years, the first core-collapse supernovae detonate, injecting enormous mechanical energy into the surrounding medium and launching a faster, more massive multiphase wind. These more evolved systems show absorption features of greater equivalent width, higher outflow velocities and higher dust content, and their Lyman continuum escape fractions are correspondingly suppressed.

The statistical support for this picture is striking. The 90th percentile outflow velocity, v90, and the equivalent width of the H I 1026 angstrom absorption line both show strong anti-correlations with the Lyman continuum escape fraction, with a Kendall rank correlation coefficient of tau equal to minus 0.87 and a p-value of 0.02. Monte Carlo resampling confirmed that these relations remain robust against measurement uncertainties. Even more remarkable, the fraction of stellar light contributed by populations between three and ten million years old shows a nearly perfect monotonic correlation with escape fraction, tau equal to 1.0 with a p-value of 0.00, although the authors caution that uncertainties in the stellar ages make the strength of this particular correlation less well constrained.

Stacked spectra of the broader LzLCS+ sample reinforce the conclusion. When Carr and colleagues combined the spectra of compact, highly star-forming galaxies split into strong leakers with escape fractions of twenty percent or more, weak leakers between five and twenty percent, and non-leakers below five percent, they found that strong leakers host the narrowest H I absorption profiles, while the C II profiles reveal that non-leakers drive the fastest outflows, followed by weak leakers. In other words, the galaxies with the most vigorous supernova-driven winds are precisely the ones leaking the fewest ionizing photons. This seems paradoxical at first glance, since winds are often invoked as the mechanism that clears escape channels. The resolution is timing: winds help open pathways, but once they launch and become heavily mass-loaded, they load the line of sight with dense, absorbing material that blocks Lyman continuum photons.

The broader implications reach to the very beginning of galaxy formation. Reionization simulations have long wrestled with the question of whether radiation pressure, stellar winds or supernovae dominate the regulation of ionizing photon escape, and theoretical models differ on when in a starburst’s life the escape fraction peaks. The new observations suggest that in compact star-forming galaxies, the critical window opens almost immediately, within the first three million years of star formation, and closes over the following several million years as supernova-driven winds launch and mass-load during the first ten million years. Because JWST is now detecting galaxies at redshifts greater than ten, whose starbursts may be only a few million years old, this timing matters enormously for interpreting whether the early universe had enough escaping photons to complete reionization, or whether additional sources such as delayed photons from binary star evolution are required.

The study also delivers practical tools for the high-redshift frontier. Indirect estimators of the Lyman continuum escape fraction, such as the far-ultraviolet continuum slope and nebular emission-line ratios, are calibrated against local measurements, and understanding how escape evolves with starburst age sharpens those calibrations. The team cross-checked their outflow velocity measurements against independent estimates derived from the [O III] 5007 line and the Mg II doublet in other surveys, finding broad agreement. All of the HST and XMM-Newton data underpinning the work are publicly available in the Mikulski Archive for Space Telescopes and the XMM-Newton Science Archive, and both analysis codes, SALT and FiCUS, are openly accessible on GitHub, inviting the community to test and extend the emerging picture. For now, the message is elegantly simple: the youngest, most violently radiative starbursts are the universe’s best window manufacturers, and their own supernova explosions are what, a few million years later, shut the window.

Subject of Research: Lyman continuum photon escape and its suppression by the onset of galactic winds in compact star-forming galaxies

Article Title: Lyman continuum escape during the onset of galactic winds in compact starbursts

Article References: Lyman continuum escape during the onset of galactic winds in compact starbursts. (n.d.). https://doi.org/10.1038/s41550-026-02999-z

Image Credits: AI Generated

DOI: 10.1038/s41550-026-02999-z

Keywords: Lyman continuum, galactic winds, starburst galaxies, reionization, Hubble Space Telescope, stellar feedback, supernovae, escape fraction, ultraviolet spectroscopy, LzLCS, interstellar medium, early universe