The claim under review
The September 14, 2015 signal recorded by the two Advanced LIGO detectors was gravitational radiation from the merger of two stellar-mass black holes.
Claim summary, not a verbatim quotation.
Our assessment
Strongly supported
What was actually measured
On September 14, 2015, the Hanford and Livingston interferometers recorded a short signal whose frequency and amplitude increased before its final decay. The discovery paper reported a network signal-to-noise ratio of 24 and a noise false-alarm rate below one event per 203,000 years. Its interpretation was a binary black-hole inspiral, merger and ringdown consistent with general relativity. These are statements about a measured transient and a tested physical model; the false-alarm rate is not the probability that the entire scientific interpretation is wrong. [1]
Why the evidence is persuasive
The companion search paper describes how the event was recovered in the search for compact-binary coalescences and how its significance was assessed against detector background. Cross-checks matter because a visually striking trace alone can be misleading: instruments produce glitches, and a search examines many candidate times and waveforms. The strength of this case is the agreement between the observations at separated sites, the expected signal structure and the noise assessment. Our methods score reflects this combination, not the fame of the result or the number of its authors. [1, 2] Specifically, the companion analysis used PyCBC and GstLAL, with different ranking statistics and background estimation procedures. Agreement between these implementations is a useful check against a single analysis pipeline creating the result, although both still operate on the same detector data. [2]
What later evidence adds
GW151226, observed on December 26, 2015 and reported in June 2016, provided another binary black-hole signal with a different mass scale and a longer observable inspiral. Its reported significance exceeded five standard deviations. This strengthened the interpretation that the instruments were observing an astrophysical population rather than one peculiar transient. It did not independently repeat the original cosmic event, and it shared the LIGO apparatus and much of the analysis infrastructure. That distinction is why replication receives four points rather than five. [3]
What outsiders can check
The Gravitational Wave Open Science Center provides strain data and supporting products for GW150914. Its event page also distinguishes the original publication from subsequently updated source-parameter estimates. Public access permits independent analysis, but access alone does not establish that every outside implementation will recover identical numbers: calibration, data selection and waveform choices still matter. The release nevertheless makes the argument unusually inspectable for an observation requiring a very large instrument. [4]
Where the conclusion stops
The review supports the detection and binary-merger interpretation. It does not certify every detail of black-hole formation, assign an exact distance without model assumptions, or prove general relativity in every regime. An event can be securely detected while its source properties remain uncertain. No claim-changing error was identified in this selected primary-source record. That is an evidence-bounded editorial finding, not an audit of all detector operations or a substitute for reproducing the analysis. The resulting grade measures support for this one published claim.
Documented problems & their consequences
No material error in this review's central claim was identified in the sources examined. That is a bounded finding, not proof that every statement in the paper is correct.
How we reached the grade
Claim support 5/5 · 40% weight
The coherent signal and its waveform support the specified astrophysical interpretation. [1, 2]
Methods & controls 5/5 · 20% weight
The search characterized detector noise and used multiple analysis approaches rather than relying on one fitted curve. [2]
Independent checks 4/5 · 20% weight
GW151226 corroborated the phenomenon, but a second event in the same instruments is not a fully independent recreation of GW150914. [3]
Transparency 5/5 · 10% weight
The collaboration released event data, parameter estimates and supporting material for outside analysis. [4]
Claim calibration 5/5 · 10% weight
The paper reported uncertainty intervals and a conditional noise significance, with a narrowly framed discovery claim. [1]
What this review does not establish
Selected primary-source review, not a fresh strain-data reanalysis. The two detectors and subsequent event share infrastructure. The grade covers the detection claim, not a lifetime assessment of collaboration members.
The evidence over time
GW150914 recorded at the two LIGO observatories.
Discovery paper published and observation announced.
GW151226 paper submitted to arXiv, adding a second binary-black-hole detection.
Primary sources & further evidence
- Discovery paper · Abbott et al., 2016 ↗
Primary detection report. Open author version: arxiv.org/abs/1602.03837.
- Search methods · LIGO and Virgo, 2016 ↗
Primary companion analysis: GW150914, first results from the binary-black-hole search; DOI 10.1103/PhysRevD.93.122003.
- GW151226 discovery · LIGO and Virgo, 2016 ↗
A subsequent event, not a rerun of GW150914. Open author version: arxiv.org/abs/1606.04855.
- GW150914 public data · GWOSC ↗
Collaboration-maintained event data and revised parameter products; dataset DOI 10.7935/K5MW2F23.
Evidence can change a grade.
Point us to a factual error, a new replication, or a missing source. Authors can submit a response through the same process.
Submit a correction or response ↗