Multi-Segment Consistency Tests of General Relativity

Abstract

As the LIGO-VIRGO-KAGRA Network of gravitational-wave detectors improves in sensitivity, accumulating hundreds of gravitational-wave detections per year, it becomes imperative to improve tests of general relativity in concert. The test of Hawking's law of area increase has gained prominence since GW250114, where black holes in General Relativity were tested with unprecedented precision, using the linear ringdown and pre-merger portions of the signal. A closely related test is the Inspiral-Merger-Ringdown Consistency Test, which assesses the consistency of the high- and low-frequency parts of the signals. In this letter, I present a multi-parameter Multi-Segment Consistency Test (MSCT) that generalizes and improves upon existing tests by ensuring that the extrinsic properties of the signal are consistent across its independent segments and by adopting an accelerated time-domain approach. The improved area law test is then presented as a projection of this MSCT test. These crucial improvements, which bring physical consistency to the area law test, lead to more stringent constraints on the increase in estimated area from observed binary black hole mergers, while also capturing covariances among the parameters. Applying the two-segment version of this test to the inspiral and ringdown parts of GW250114, and keeping some of the extrinsic parameters common between the segments, I test the signal to unprecedented accuracy, obtaining $4.61 ^{+0.24} _{-0.11}σ$ significant result for the area increase, even as more than 4 pre-merger cycles of the signal are excluded from the analysis. Also, I infer that the final state lies within the 15\% highest posterior density confidence interval.

Figures (6)

• Figure 1: The fractional area increase from the MSCT, carried out on GW250114. The blue histogram shows the area increase from the multi-segment method, while that in orange shows the expected GR area increase from the posterior samples obtained from the full IMR PE.
• Figure 2: The ratio of the logarithms of the area increase predicted by the MSCT to that from full IMR analysis. The orange-dotted lines indicate the 10% and 90% quantiles. The median is indicated in cyan, and the GR value, i.e., 0, is represented by the red solid line.
• Figure 3: Consistency of the (semi-)independent predictions for the initial and final areas from each segment. The GR value is well within the 90% confidence interval.
• Figure 4: The consistency of the marginal distributions of four of the 10 exclusive parameters across the segments. The bounds are presented in Tab. \ref{['tab:diff_pars_and_pars']}
• Figure 5: Consistency in the ringdown parameters. The left panel shows the 50, 68, 90, and 95% configdence contours for the predicted final mass and spin in the I, R, and IMR analysis modes. The red cross indicates the maximum-likelihood waveform from the full IMR time-domain analysis. The right panel shows the difference between the predictions from the two segments. The GR value 0 lies at at boundary of the $15\%$ CI regions, which corresponds to $\approx 0.2 \sigma$