A Sign Flip in a Galaxy-Rotation Diagnostic, and an Open Platform Built to Check It

Astronomers cannot watch a single galaxy spin up over billions of years. They can, however, line up snapshots from different cosmic epochs and ask whether the rules of rotation stay the same. A new preprint by David C. Flynn, released on arXiv, describes an open-science platform built to make that comparison possible—and highlights one number that appears to flip sign between the nearby universe and the distant past.

The number is omega, a boundary-point angular velocity. It is derived from the innermost and outermost resolved rings of a galaxy's rotation curve, the plot of orbital speed versus distance from the center. In the local-universe samples, omega is positive on average. In a small sample of star-forming galaxies from the early universe, it is negative. The paper reports a sign reversal of about 20 radians per billion years (rad/Gyr) across roughly 9 billion years of cosmic evolution. The sign of omega is not a simple spin direction; it is a diagnostic of how the rotation signal changes from the inner to the outer galaxy, and it may be tracing the structural transformation from compact high-redshift systems to extended disks.

The Radio and Millimeter Traces of Rotating Gas

To measure rotation across cosmic time, astronomers use different tracers. In the local universe, the 21-centimeter line of neutral hydrogen (HI) maps gas disks in and around galaxies. At high redshift, where galaxies are often compact and centrally concentrated, ALMA observes the [CII] line from singly ionized carbon, a line used for morpho-kinematics in star-forming galaxies. Redshift, denoted z, measures how much cosmic expansion has stretched a galaxy's light; z=0 is today, while the platform reaches to z~6, approaching the epoch of reionization. The preprint's high-redshift sample sits at z=4.26-5.68.

The new platform brings these tracers under one schema. It includes four machine-readable corpora totaling 772 objects: 438 HI-traced galaxies from SPARC, THINGS, LITTLE THINGS, and WALLABY DR2; 129 dwarf and irregular galaxies; 174 Milky Way globular clusters; and 31 ALPINE star-forming galaxies at high redshift. The data are released in JSON, flat CSV, RAG-ready JSONL, and per-object ZIP files, with permanent Zenodo DOIs under a CC BY 4.0 license.

A Number That Changes Sign

The scientific centerpiece is the omega correction developed by Flynn and Cannaliato in 2025. In this preprint, Flynn applies it to the platform's rotation-curve samples. For 84 SPARC galaxies with the original Q=1 quality flag, the mean omega is +7.06 ± 3.26 rad/Gyr at z=0. A set of 24 dwarf galaxies meeting the omega-ready criterion has a median of +9.94 rad/Gyr. At high redshift, the picture changes: eight confirmed rotators in the ALPINE sample all have negative omega, with a median of -13.05 rad/Gyr. The paper calls this a sign reversal of about 20 rad/Gyr across roughly 9 billion years.

The author suggests the flip is consistent with a structural transformation. High-redshift star-forming galaxies are often compact and centrally concentrated, while nearby disk galaxies are extended rotating disks. The omega parameter may be capturing that structural change. But the paper does not claim to have proven the cause. It motivates future RAMSES cosmological simulations, planned for 2026-2027, to trace how such galaxies evolve from z~6 initial conditions.

The Small Sample Behind the Signal

The high-redshift evidence rests on a small number of objects. The Z1 corpus contains 31 galaxies, but only eight are confirmed rotators with per-ring rotation curves; the rest have morpho-kinematic classifications only. That means the negative omega median comes from eight confirmed rotators. The platform paper itself is an arXiv preprint. The omega framework was previously published, but its application to this unified cross-epoch sample is presented as new.

Even the term sign reversal needs care. The sign of omega is not a simple spin direction: it is a parameter derived from the rotation curve's inner and outer rings, and its sign depends on how those rings are compared. The paper's claim is that the sign of this specific diagnostic changes systematically between the local and high-redshift samples.

Open Data, Notebooks, and a High-School Track

The platform is not only a scientific result. It is an attempt to lower barriers. It provides 120 Jupyter notebooks, all verified to run without errors, covering rotation-curve plotting, baryonic decomposition, globular cluster dynamics, and high-redshift [CII] analysis. A QuickStart notebook reproduces the cross-epoch omega comparison in under 10 minutes from a fresh clone. A High-School Exploration Track offers 20 notebooks split by age group, using the same published data rather than simplified stand-ins. No telescope archive account, proprietary software, or institutional access is required.

A Stage 2 planned for 2026 would add fine-tuned large language models to help query and extract information from the corpora, with adapters ranging from research-cluster scale to laptop-compatible sizes. That part is a roadmap, not a delivered product. The author also discloses that four large language models were used during development and validation, with conclusions remaining the author's responsibility.

Checking the Omega Sign Reversal with Open Data

The omega sign reversal is the kind of result that can be checked, challenged, and extended because the data and code are openly available. If it holds, it would add a new cross-epoch clue to how galaxies settle into rotating disks. If it does not, the platform still gives researchers and students a chance to work with a common kinematic framework spanning from the Milky Way's globular clusters to galaxies approaching the epoch of reionization. Either way, the story is now one of open data: a sign flip in a single parameter, openly available for checking and extension.