Euclid DR1 Spectroscopy Agrees with DESI Redshifts for 98.5–98.9% of High‑Quality Sample
A new Euclid Collaboration paper says the European Space Agency mission’s first major public spectroscopic dataset matches redshift measurements from the Dark Energy Spectroscopic Instrument, or DESI, at 98.5% for a high-quality comparison sample in the Euclid Deep Field North. With tighter pipeline confidence cuts, the agreement rises to 98.9%.
That benchmark matters because Euclid is opening a vast public spectroscopy dataset to astronomers, and redshift measurements — which show how far away galaxies are and help place them in cosmic history — are central to the mission’s 3D map of the universe. The results are an early sign that Euclid’s space-based spectroscopy pipeline is performing well on selected, high-quality samples. They are not a claim of overall mission-wide accuracy for every spectrum in the release.
The findings are described in the paper “Euclid Data Release 1 (DR1): Performance of the redshift measurement and spectroscopic analysis,” posted to arXiv as arXiv:2611.17960. The study focuses on Euclid’s DR1 spectroscopic dataset and the automated processing system known as the Spectroscopic Processing Function, or SPE PF, which analyzes 1D spectra from the mission’s Near-Infrared Spectro-Photometer, or NISP. The pipeline handles spectroscopic classification, redshift determination and measurements of spectral features.
To test the results, the team cross-matched Euclid’s measurements with DESI DR1 and also carried out a visual-inspection campaign covering about 4,000 sources in the Euclid Deep Field North, one of the mission’s repeatedly observed calibration fields. For a high-quality EDF-N sample, Euclid and DESI spectroscopic redshifts agreed for 3,985 objects at the 98.5% level. With additional pipeline-based confidence cuts, using a sample of 3,863 objects, the agreement increased to 98.9%.
Euclid was launched on July 1, 2023, and operates from the Sun-Earth L2 point, with routine science observations beginning Feb. 14, 2024. Its core cosmology program relies on two main tools: weak gravitational lensing, which traces matter through subtle distortions in galaxy shapes, and galaxy clustering, which maps how galaxies are distributed through space and time. The spectroscopy in this paper underpins the second of those goals.
NISP uses slitless near-infrared spectroscopy, meaning many objects in the same field produce dispersed spectra at once rather than being isolated behind individual slits. That approach is efficient for a large sky survey, but it also creates well-known problems, including overlapping spectra, contamination from nearby sources and possible line misidentifications. The paper says the team mitigates those issues with pixel flags, valid-pixel cuts and priors. For Euclid’s cosmology program, NISP uses a red grism with a spectral resolution of about 500, covering roughly 1206 to 1892 nanometers. Euclid’s target redshift range for H-alpha-based galaxy clustering is 0.84 to 1.88. The pipeline measures emission-line fluxes in two independent ways — direct integration and Gaussian fitting — and switches to a dedicated “QSO mode” for sources with broad H-alpha emission above a reported velocity-dispersion threshold of 1,500 kilometers per second.
The authors say the DR1 spectroscopic dataset provides reliable measurements for millions of extragalactic sources and establishes a baseline for large-scale structure and galaxy-evolution studies. The release is part of DR1-Foundation, Euclid’s first major public data tranche, issued Nov. 12, 2026. ESA says it covers about 13 months of observations and includes spectra from the three Euclid Deep Fields and about 1,900 square degrees of the Euclid Wide Survey, totaling more than 160 million spectra alongside imaging of more than 900 million galaxies. As ESA’s Euclid project scientist Valeria Pettorino said in the agency’s DR1-Foundation release, “This is a change in scale from Euclid’s previous releases, which included a few days of observations over around 60 degrees squared of sky.”