SPHEREx shell around VY Canis Majoris identified as image-processing artifact in new arXiv analysis
The preprint demonstrates that the reported SPHEREx shell is an artifact of background subtraction aligned with scan direction. Re-reduction with updated pipelines eliminates the feature while preserving known mass-loss structures. This underscores the need for rigorous artifact testing in infrared survey data before interpreting extended emission around bright stars.
The analysis reprocesses early SPHEREx all-sky survey frames centered on the red hypergiant using independent flat-fielding and stray-light models. After applying updated zodiacal-light templates and testing multiple dither combinations, the apparent 10-arcminute shell drops below 3-sigma significance in all tested bands. Residual maps show the feature spatially correlates with the spacecraft's scan-direction artifacts rather than with the star's known asymmetric mass-loss history.
Prior claims relied on a single-pass pipeline that left low-level gradients aligned with the survey's scan pattern. Re-reduction with an independent team pipeline and cross-checks against WISE 12-micron and AKARI far-infrared maps find no corresponding emission at the reported surface brightness. This pattern matches earlier cases where uncorrected 1/f noise in infrared surveys produced false extended structures around bright point sources.
If confirmed, the result tightens requirements for artifact rejection in SPHEREx's public data releases and similar all-sky infrared missions. It also preserves the existing multi-wavelength picture of VY CMa mass loss derived from ALMA, HST, and ground-based adaptive-optics imaging without requiring an additional large-scale shell component.
Independent verification is expected once the next SPHEREx data release includes improved calibration products; observers can test the claim by stacking the same fields with the forthcoming pipeline version.
SPHEREx team: independent re-reduction of the same fields will show the shell feature remains below detection threshold in the next public data release within 9 months.
Sources (2)
- [1]Primary Source(https://arxiv.org/abs/2609.28612)
- [2]Supporting Source(https://ui.adsabs.harvard.edu/abs/2024ApJ...961..145H)