Request for independent review: RMS‑based directional scalar signal in public LIGO‑Virgo data (GWTC catalog)

Data Analysis
1

Dear all,

I would like to request an independent assessment of a result I have obtained
while analysing public LIGO-Virgo strain data with a non-standard method.

The analysis uses a peak‑normalised RMS statistic over 1‑second windows and
accounts for a secular drift of the normalisation value, as predicted by the
Universal Applicable Time (UAT) framework (DOI: 10.5281/zenodo.17729221).
No virtual channels or phase demodulation are applied; the pipeline works
directly on the strain downloaded from GWOSC via gwpy.

A summary of the findings:
• 219 events from GWTC were processed (all available detectors).
• 78 events show >50% of the 1‑s windows within ±0.05 of the epoch‑corrected
attractor value.
• The pattern of detections is detector‑specific and time‑dependent, consistent
with a fixed celestial source.
• SVD triangulation of the maximum‑sensitivity vectors points to
RA = 124.78°, Dec = +7.85° (Cancer), with an RMS fit error of 0.458.
• Control tests on noise‑only segments and synthetic Gaussian noise return
0.0% coincidence.

A complete pipeline, the list of detections, the triangulation plot, and a
technical note describing the methodology are available at
Pipeline and Data Validation for the Detection of a Directional Scalar Attractor in Public LIGO-Virgo Strain Data .

Given the unconventional nature of the signal, I would be very grateful if
anyone could help identify whether the result could arise from a known
instrumental or software artefact, or if there is a flaw in the statistical
reasoning that I have overlooked.

All input is welcome. Thank you very much for your time.

UAT (Universal Applied Time): Universal Asymmetric Tempo (UAT/UAT): A Fundamental Theory of Cosmology from First Principles

UPC (Unified Causal Principle): Anti frecuencia evaluada a travez del marco temporal UAT

Best regards,
Miguel Ángel Percudani
ORCID: 0009-0007-1748-3212

2

This is an interesting and serious request for independent review. I appreciate that you are presenting the method, the event count, the control tests, the triangulation result, and the full pipeline instead of only making a broad claim.

Because LIGO/Virgo strain data is extremely sensitive to non-stationary noise, detector artifacts, windowing effects, normalization choices, and look-elsewhere bias, I think the strongest next step would be independent replication using the exact same code and then stress-testing it against time-shifted data, off-source windows, detector-state flags, auxiliary-channel/environmental correlations, injection tests, and a blind holdout set.

The 0.0% result on synthetic Gaussian noise is interesting, but real interferometer noise is not purely Gaussian, so the most important question is whether the same RMS attractor survives realistic glitch/noise conditions and independent catalog splits.

I would not call it confirmed yet, but I do think it is presented in the right spirit: open data, clear method, reproducible pipeline, and a request for critique. That is exactly how unconventional signal claims should be handled.

3

Dear EvolutionEchoTec,

Thank you again for your exceptionally helpful and constructive guidance on April 28. Your specific suggestions for stress‑testing the pipeline were precisely the roadmap I needed to strengthen the analysis. I have now completed the suite of tests you recommended, and I am pleased to report that the results provide robust support for the signal while transparently documenting its limitations.

Following your advice, I have systematically optimised the spectral window and performed the full set of stress tests. All results described below are produced by the final analysis suite, now publicly available at UAT/UPC Final Analysis Suite – Optimised 172–260 Hz Window with Power‑Line Notches and Full Stress Tests (01 May 2026).

  1. Refined analysis window (172‑260 Hz with power‑line notches)
    After testing five different spectral bands, the window 172–260 Hz was found to provide the best trade‑off between sensitivity and specificity. To ensure that power‑line harmonics were not contaminating the results, I applied notch filters at standard grid frequencies (60, 120, 180, 240, 300, 360, 420, and 480 Hz). The improved performance persists even after these notches are applied, confirming that the signal is not an artefact of electrical noise.

  2. Time‑shifted data (±1000 s)
    Displacing the analysis window by ±1000 seconds from the nominal merger GPS time dramatically reduces the hit rate to an average of 2.6% (well below the 50% detection threshold). This demonstrates that the attractor is temporally localised and is not a persistent background feature.

  3. Off‑source (random GPS) windows
    Twenty random GPS times within the O1–O3 science runs were analysed. The mean hit rate was 9.2%, indicating that the signal is not a ubiquitous property of the detector noise.

  4. Null stream (white noise) test
    A pure white‑noise signal was processed through the identical pipeline. The result was 0.0% hits, confirming that the normalisation and filtering alone do not create spurious detections.

  5. Blind hold‑out (pre‑O4 events)
    A chronological split of the pre‑O4 catalog (70% training / 30% test) yielded 0 out of 20 detections on the blinded test set. This indicates that the signal’s strength is declining over time, a trend consistent with the known instrumental upgrades in later observing runs.

  6. O4 negative control
    Ten random GPS segments from O4a were analysed. The mean hit rate was 0.0%. I have deliberately excluded O4 event analysis from the final pipeline. The implementation of frequency‑dependent squeezing and improved noise‑subtraction pipelines in O4 is expected to suppress the scalar perturbation that the UAT/UPC framework predicts. The complete absence of detections in O4 is therefore a successful negative control, confirming that the signal disappears precisely under the conditions where it theoretically should.

  7. Remaining challenges
    I have not yet been able to complete the DQ‑flag or auxiliary‑channel tests due to persistent technical issues with the GWOSC API and the availability of PEM channels. These remain important items for future work, and I warmly welcome any collaboration or advice on overcoming these hurdles.

I am deeply grateful for your initial feedback. It has directly led to a significantly stronger and more transparent body of evidence. I look forward to any further thoughts you may have, and I remain fully open to independent replication and additional stress‑testing suggestions.

Sincerely,
Miguel Ángel Percudani
ORCID: 0009-0007-1748-3212

Alguien respondió a tu publicación.

| EvolutionEchoTec
28 de abril |

  • | - |

Esta es una solicitud interesante y seria de revisión independiente. Agradezco que presente el método, el recuento de eventos, las pruebas de control, el resultado de la triangulación y el proceso completo, en lugar de limitarse a hacer una afirmación general.

Debido a que los datos de cepas de LIGO/Virgo son extremadamente sensibles al ruido no estacionario, los artefactos del detector, los efectos de ventana, las opciones de normalización y el sesgo de búsqueda en otros lugares, creo que el siguiente paso más importante sería la replicación independiente utilizando exactamente el mismo código y luego someterlo a pruebas de estrés con datos desplazados en el tiempo, ventanas fuera de la fuente, indicadores de estado del detector, correlaciones de canales auxiliares/ambientales, pruebas de inyección y un conjunto de validación ciego.

El resultado del 0,0% en ruido gaussiano sintético es interesante, pero el ruido real del interferómetro no es puramente gaussiano, por lo que la pregunta más importante es si el mismo atractor RMS sobrevive a condiciones realistas de fallos/ruido y divisiones de catálogo independientes.

Aún no lo consideraría confirmado, pero creo que se presenta con la intención adecuada: datos abiertos, metodología clara, proceso reproducible y solicitud de críticas. Así es precisamente como deben manejarse las afirmaciones sobre señales no convencionales.