Looking for ancillary signals around GW150914

Abstract

We replicated the procedure in Liu and Jackson [1], who had found evidence for a low amplitude signal in the vicinity of GW150914. This was based upon the large correlation between the time integral of the Pearson cross-correlation coefficient in the off-source region of GW150914, and the Pearson cross-correlation in a narrow window around GW150914, for the same time lag between the two LIGO detectors as the gravitational wave signal. Our results mostly agree with those in Liu and Jackson [1]. We find the statistical significance of the observed cross-correlation to be about 2.5 σ. We also used the cross-correlation method to search for short duration signals at all other physical values of the time lag, within this 4096 second time interval, but do not find evidence for any statistically significant events in the off-source region.

Pearson cross-correlation in the first four black hole binary mergers

“…Monitoring the correlation in the data collected by independent interferometers constitutes an interesting method to test the unexpected. Some publications have explored this direction analyzing the LIGO data streams with known events [16][17][18][19][20][21]. It turns out that several subtleties jeopardize the reliability of such a test [22].…”

“…[19] concludes that this mismatch for GW150914 is not statistically significant once one takes care of subtleties, such as whitening versus notching or discrepancies in the best fit waveforms. Still, some debate remains over the exact size of the statistical significance [20,21].…”