Water ice is the main reservoir of oxygen and a building material for planets in star formation regions. Its abundance and distribution in prestellar cores dictate the chemistry of future protoplanetary disks. Until recently, ice maps were pieced together from sparse lines of sight toward background stars, yielding a fragmentary view. Launched in 2025, the SPHEREx telescope conducts a spectrophotometric all-sky survey in 102 narrow bands, making it possible to use the scattered infrared light from the clouds themselves—coreshine—for continuous ice mapping.
For four nearby (∼140 pc) prestellar cores of varying density, spectra were extracted from SPHEREx data (channels D4 and D5, 2,4–4,4 μm) on a grid with 7,2 arcsec spacing (∼1000 AU). After removing pixels contaminated by starlight or with low signal-to-noise, a continuum was fitted with a quadratic function using regions outside the ice band (2,4–2,7 and 3,6–4,1 μm). The optical depth of water ice τ was calculated as the average logarithm of the flux deviation from the continuum over the 2,9–3,1 μm range. In parallel, analytic and three-dimensional Monte Carlo models (RADMC-3D) of Bonnor–Ebert spheres with realistic scattering and absorption by dust with icy mantles were constructed.
In cores L1512 and L260, with central densities of ∼10⁴–10⁵ cm⁻³, ice absorption increases monotonically inward, reaching τ≈0,8–1. In contrast, the ultra-dense cores CrA 151 and L1544 (∼10⁷ cm⁻³) show a dip: τ drops in the very center, where it was expected to peak. Models fail to reproduce this—changing the illumination geometry, growing grains up to 100 μm, or imposing an ice formation threshold all fail to produce a central minimum. Only scenarios involving saturation or a sharp drop in the water ice fraction in the densest core qualitatively match the observations. A comparison with literature data indicates that along lines of sight with AV>25–30 mag, the correlation between ice abundance and density begins to flatten.
The results demonstrate that spectroscopy of coreshine is a powerful tool for ice mapping, independent of the random placement of background stars. The central dip uncovered in ultra-dense cores points to a gap in our understanding of ice evolution: either water ice is converted into other species (like CO ice or organics), or physical conditions in core centers alter scattering efficiencies or absorption coefficients. The method is applicable to all cores with detectable coreshine—roughly half of known cores.
With large samples of cores with ice maps from SPHEREx, future studies will statistically probe relationships between ice distribution, core evolutionary stage, and environment. Refining the models will require incorporating radiation field anisotropy and more realistic dust and ice compositions. Joint analysis of H₂O, CO, and CO₂ bands in SPHEREx spectra will recover the complete oxygen budget and test hypotheses about chemical evolution in dense cores.
The method dramatically expands the reach of observational astrochemistry for prestellar cores and protoplanetary environments, enabling ice studies at the scales of forming planetary systems. The findings will influence models of the initial chemical composition of protoplanetary disks and, ultimately, our understanding of water and organic delivery to planets.
Plans include extending the analysis to dozens of cores with known coreshine morphology and incorporating more detailed microphysics of dust and ice into models. A key goal is directly comparing maps of water ice with those of other icy species (CO, CO₂) from SPHEREx data.
The central absorption dip of water ice in the densest cores directly addresses the unsolved problem of interstellar ice chemical evolution: where does oxygen go in pre-collapse conditions? This may connect to the mystery of oxygen depletion in the interstellar medium or to overlooked channels for complex organic formation.
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
For the first time, optical depth maps of water ice in four prestellar cores have been constructed using SPHEREx data, based on the scattered light of the clouds themselves (coreshine). In two superdense cores, a central absorption dip was found, contradicting model predictions.
Where its strength is
The ice mapping method using coreshine does not depend on the random positions of background stars and provides continuous coverage, opening the possibility of systematic studies of ice evolution in cores.
Where it could go next
Worth trying
Why this may matter
If the central dip is indeed related to the chemical evolution of ice, it will change our understanding of the oxygen balance and the composition of building blocks of planetary systems.
Nearby works in our archive
The closest work in the archive sits at 0.65 by meaning — the usual level for our corpus.
SPHEREx is a real cosmic 'ice cartographer': every six months it maps the entire sky, and its data already contain millions of spectra, ripe for finding frozen water's fingerprints in the cradles of future planets.
Although diffuse glow from prestellar cores hasn't directly featured in science fiction plots, the idea of 'history-keeping ice' echoes the giant ice library in Kim Stanley Robinson's novel '2312'—only here, it stores the chemical evolution of a cloud.