Residual Test for the Third Gravitational-Wave Transient Catalog

Abstract

The residual test is commonly used to check the agreement between the gravitational wave signal and the theoretical waveform template. The basic idea of the residual test is to subtract the best-fit waveform from the data and then check whether the remaining data (i.e., the residuals) are consistent with the instrumental noise or not. We apply the Kolmogorov-Smirnov test, the Anderson-Darling test and the chi-squared test as goodness-of-fit test to examine the residuals of events in the third gravitational-wave transient catalog and find no statistically significant deviation from the noise. Although our method is sensitive only to the loud events, it does not rely on the cross-correlation between detectors. A single-detector event suffices for our residual analysis, and the test is simple and computationally inexpensive.