Preprint claims first detection of 'moving lens' effect in CMB at 4.8σ

A newly posted arXiv preprint reports what its authors describe as the first detection of the “moving lens effect,” a subtle distortion in the cosmic microwave background, or CMB, with a top statistical significance of 4.8 sigma. If confirmed, the result could give cosmologists a new way to measure how matter moves sideways across the sky, not just toward or away from Earth.

The moving lens effect is a tiny secondary imprint on the CMB — the afterglow of the early universe — caused by large concentrations of matter moving across our line of sight. In plain terms, the gravity of those structures slightly alters the CMB signal as they sweep sideways. That makes the effect different from the better-known kinematic Sunyaev-Zel’dovich, or kSZ, effect, which is used to study motion along the line of sight. The new paper argues that the moving lens signal could open direct access to transverse cosmic velocities.

The paper, titled “First detection of the moving lens effect with ACT and DESI LS,” was posted to arXiv on May 18 as version 1 of preprint 2605.18938. Its authors are Selim C. Hotinli, Kendrick M. Smith, Simone Ferraro, Ali Beheshti, Arthur Kosowsky, Elena Pierpaoli and Emmanuel Schaan. The analysis combines CMB temperature maps from Atacama Cosmology Telescope Data Release 6, a major recent CMB data set, with luminous red galaxies from the DESI Legacy Imaging Surveys, a wide-field imaging program that provides galaxy catalogs tracing large-scale structure.

The main reported result comes from the foreground-reduced ACT NILC map. There, the authors report a moving-lens amplitude of ( b_{\rm ML} = 1.24 \pm 0.26 ) for an extended galaxy sample, corresponding to 4.8 sigma significance. For a main sample, they report ( b_{\rm ML} = 0.93 \pm 0.25 ), or 3.7 sigma. The paper says both measurements are consistent with halo-model predictions for the moving-lens signal.

To make the measurement, the authors say they developed a Fourier-space cross-spectrum estimator that preserves the signal’s dependence on scale. They also say their analysis pipeline enforces a separation of scales between the reconstructed velocities and the cross-correlation step, which they argue is important for reducing contamination from astrophysical foregrounds. On possible systematics, the preprint says residual foreground contamination is expected to be significantly smaller than the signal, that no curl-mode test exceeds 2 sigma, and that the results remain robust across analysis variants.

The claim builds on several years of theoretical and forecasting work, including papers from 2018 onward, that argued modern CMB maps and galaxy surveys might finally be sensitive enough to detect the effect. According to the research brief, no earlier published observational detection had been identified before this preprint.

Why that matters is straightforward: cosmologists already have tools such as the kSZ effect to probe line-of-sight motion, but sideways motion has been much harder to access directly. The authors argue that combining the moving lens effect with kSZ measurements could eventually help map the universe’s three-dimensional velocity field.

For now, though, the result remains a first-detection claim in a new arXiv preprint, not a peer-reviewed journal paper. Whether the measurement holds up will be assessed through peer review and follow-up analyses using the same and future survey data.