Gravitational Waves - Are all detectors finding the same gravitational waves? I read that there have been approximately 90 recorded cases of gravitational waves. Have the 4 different gravitational wave detectors agreed on specific individual recordings or have all 90 cases been identified by individual detectors but not confirmed by other detectors?
 A: The most recent catalogue of GW sources includes detections made by the LIGO + VIRGO instruments. Thus up to three detectors were working at any one time.
Most of the signals have been corroborated with almost-simultaneous detections (remember that GWs travel at the speed of light and the detectors are separated by thousands of km) in more than one detector. However, some would not be considered significant in one or sometimes in any of the detectors individually, but in total they combine to give a significant detection with a very small chance they are of non-astrophysical origin.
For example you can look at table IX in Abbott et al. (2021), which gives the signal to noise ratios seen in each of the individual detectors. There you can see for example that GW200311_103121 was observed by Hanford and Livingston (the LIGO interferometers) but not detected with high significance (a SNR$<8$) in either of the individual observations.
There is also a single claimed GW event in run O3a (GW190424_180648) which was only observed with the Livingston detector (see Abbott et al. 2020), although it is only given a 91% probability of being of astrophysical origin.
A: Most, but not all, detections have been detected simultaneously$^\star$ by two or more detectors. It helps to know a little history.

*

*O1 (observing run 1): September 2015 - January 2016. LIGO-Hanford and LIGO-Livingston were both operating, and all events were detected by both detectors. (I think there were 3 total events detected, where one of these was first announced as a candidate and later upgraded to a detection). In fact, the consistency of the signal in both detectors (in terms of arrival time and other parameters) is an important criteria for being able to claim a detection at all, because it reduces the probability of a terrestrial effect causing the signal. (It's much harder to think of terrestrial effects that can cause two gravitational wave detectors to go off simultaneously, than to think of effects that would cause a trigger in one detector).


*O2: December 2016 - August 2017: LIGO-Hanford and LIGO-Livingston were operating the entire time and both detected all (approximately 10) events. Virgo was added to the network in August, and participated in the detection of at least 2 events. Ironically, Virgo did not detect GW170817 (the famous binary neutron-star multi-messenger event), but the non-detection of Virgo actually helped to constrain the sky position of the event significantly (because, essentially, the event had to be in a blind spot for Virgo), which was crucial in enabling the multi-messenger detection.


*O3: April 2019 - March 2020: LIGO-Hanford, LIGO-Livingston, and Virgo were all operating. KAGRA also started running, but not at a sufficient level of sensitivity to detect events. There were a mix of events that were detected by all three detectors, by two detectors, and for the first time, "single" events detected by only one detector were discovered. (Detecting "singles" required a lot of development of data analysis algorithms, and probably would not have been possible as a first discovery of GWs, but became possible as there was increased confidence working with data). The main factor determining the number of detectors seeing an event is simply which combination of detectors is online and in observing mode at the time the event occurs. There is regular maintenance performed where one or more of the detectors is taken offline, and seismic events can lead to one or more of the interferometers losing lock, meaning they are not able to observe until they are re-locked.


*O4 is supposed to start later this year, if all goes well.

$^\star$ By "simultaneously", I mean "up to the light travel time between the detectors (which is order 10 ms for Hanford and Livingston)." The small time delay (as well as phase shifts and differences in the amplitude of the signal due to the interaction of the polarization of the wave and the detector antenna pattern) allow the network to localize the events on the sky.
