Galaxy motions shift a megamaser estimate of cosmic expansion downward

Six galaxies offer a way to measure cosmic expansion without relying on the traditional astronomical distance ladder. But their movements through their surroundings complicate the measurement. A reanalysis by Richard Watkins of Willamette University and Hume A. Feldman of the University of Kansas finds that accounting differently for those motions lowers the expansion rate inferred from the sample.

Their preferred estimate of the Hubble constant is 69.7 ± 2.7 kilometres per second per megaparsec. This quantity describes how rapidly recession speed increases with distance; a megaparsec is about 3.26 million light-years. The result is closer to the lower value inferred from the cosmic microwave background than to the higher distance-ladder estimate. Crucially, however, its uncertainty remains large enough to accommodate both.

The work is an arXiv preprint, first submitted in August 2026 and revised on November 6. The supplied landing page lists no journal publication or acceptance information. Its conclusions should therefore be presented as findings from a preprint, not as an established resolution of the expansion-rate dispute.

Why these galaxies matter

The disagreement known as the Hubble tension concerns two precise but different estimates. The paper quotes 67.4 ± 0.5 kilometres per second per megaparsec from the cosmic microwave background, interpreted through the standard cosmological model, and 73.5 ± 0.81 from the local distance ladder.

The ladder connects successive distance-measuring techniques to reach progressively farther objects. Megamasers provide a different route: exceptionally bright microwave emission associated with gas orbiting galactic central black holes allows astronomers to determine distances geometrically. These distances can then be compared with the host galaxies’ redshifts to estimate cosmic expansion.

Watkins and Feldman reuse distance and redshift measurements for six galaxies from a 2020 study led by Dominic Pesce. They do not report new megamaser observations. All six galaxies lie at redshifts below 0.04, where a galaxy’s motion relative to the overall expansion can make a consequential contribution to its observed redshift.

Separating expansion from local motion

These additional movements are called peculiar velocities. Gravity can pull galaxies together into coherent flows, while individual galaxies can also have unusually large motions. Treating those velocities simply as independent, bell-curve-distributed noise can miss both effects, the authors argue—especially when the sample contains only six objects.

Their main correction uses a reconstruction of nearby galaxy motions developed by McAlpine and colleagues. It combines the 2M++ galaxy redshift survey with cosmological simulations constrained to represent the local universe. The reconstruction supplies 80 plausible versions of the velocity field rather than one definitive map.

For each version, Watkins and Feldman locate the megamaser hosts using their observed redshifts while accounting for the reconstructed motions. Where several positions match an observed redshift, they average over those solutions. They then correct the redshifts and calculate how well different Hubble constants fit the measured distances.

Importantly, they average the resulting likelihoods across the 80 reconstructions, rather than first averaging each galaxy’s velocity. This retains the linked pattern of motions within each reconstructed universe. Their preferred analysis also assigns each galaxy a separate velocity uncertainty based on the variation among reconstructions.

A lower estimate, not a verdict

In the new reconstruction, all six hosts have positive mean peculiar velocities: their local movements add to their apparent recession. Subtracting that contribution reduces the expansion rate inferred from their distances. The three farthest galaxies also appear to share a large coherent outflow, meaning their motions cannot safely be expected to cancel one another.

The authors’ comparison analysis without explicit velocity corrections gives 73.6 ± 3.0 kilometres per second per megaparsec. Their preferred corrected result is roughly four units lower. Applying an older reconstruction with the positioning procedure intended by its developers also lowers the estimate, to 70.2 ± 2.6.

Several limitations prevent a decisive conclusion. The sample is small, distance uncertainties vary widely, and the assigned velocity errors affect how strongly each galaxy influences the answer. The inferred motions also depend on how galaxies are positioned within the reconstruction. Although the distances bypass the traditional ladder, the overall analysis is not assumption-free: it uses simulation-based velocity models and fits a flat cosmological model with a fixed matter-density parameter.

The authors suggest that an unidentified distance-ladder systematic deserves continued scrutiny. This analysis does not identify such an error or rule out new physics. Its firmer lesson is methodological: nearby galaxy motions can materially shift a megamaser expansion estimate. More megamasers distributed across a broader region of sky, better distances and improved velocity reconstructions are needed to make this independent test decisive.