Two very loud gravitational-wave signals from merging black holes were used to check how well the detectors are calibrated. The events, called GW240925 and GW250207, were seen by the LIGO Hanford, LIGO Livingston and Virgo network with network signal-to-noise ratios of about 32 and 69 respectively. Signal-to-noise ratio (SNR) is a measure of how strong the signal is compared with the detector noise.

Gravitational waves from coalescing binaries have a predictable pattern of changing phase and amplitude that comes from general relativity. Normally researchers combine those predicted waveforms with independent measurements of the instruments’ calibration to infer the source properties. Calibration here means the mapping from the electrical output of a detector to the true strain in spacetime that the wave produced.

For very loud signals, the waves themselves can provide information about calibration. The authors present the first informative ‘‘astrophysical calibration’’ measurements, meaning they used the observed signals to constrain the detectors’ calibration. For GW240925 they checked the calibration inferred from the astrophysical signal against known calibration errors measured in situ. In-situ measurements are the usual engineering checks done on the instruments at the observatory.

The two events behaved differently in important ways. At the time of GW250207 the Hanford detector was not fully stabilized, which led to larger calibration uncertainties. In that case using the astrophysical signal to help calibrate the detector became essential for getting accurate data and for locating the source on the sky.

This approach matters because well-localized, high-SNR observations can give more precise measurements of source properties, enable stricter tests of general relativity, and improve so-called dark siren measurements (using gravitational-wave events without a detected light signal to learn about the Universe). All of these opportunities depend on properly accounting for calibration uncertainties.