IGEC2: A 17-month search for gravitational wave bursts in 2005-2007

TL;DR

This paper reports on a 17-month search for gravitational wave bursts using four resonant mass detectors, achieving high duty cycles and low false alarm rates, but finding no gravitational wave candidates.

Contribution

First long-term multi-detector search for gravitational wave bursts with resonant mass detectors, establishing a methodology and setting upper limits on event rates.

Findings

No gravitational wave candidates detected.

Achieved a false alarm rate of 1 per century.

Maintained high duty cycle with 57% with all four detectors.

Abstract

We present here the results of a 515 days long run of the IGEC2 observatory, consisting of the four resonant mass detectors ALLEGRO, AURIGA, EXPLORER and NAUTILUS. The reported results are related to the fourfold observation time from Nov. 6 2005 until Apr. 14 2007, when Allegro ceased its operation. This period overlapped with the first long term observations performed by the LIGO interferometric detectors. The IGEC observations aim at the identification of gravitational wave candidates with high confidence, keeping the false alarm rate at the level of 1 per century, and high duty cycle, namely 57% with all four sites and 94% with at least three sites in simultaneous observation. The network data analysis is based on time coincidence searches over at least three detectors: the four 3-fold searches and the 4-fold one are combined in a logical OR. We exchanged data with the usual blind procedure, by applying a unique confidential time offset to the events in each set of data. The accidental background was investigated by performing sets of 10^8 coincidence analyses per each detector configuration on off-source data, obtained by shifting the time series of each detector. The thresholds of the five searches were tuned so as to control the overall false alarm rate to 1/century. When the confidential time shifts was disclosed, no gravitational wave candidate was found in the on-source data. As an additional output of this search, we make available to other observatories the list of triple coincidence found below search thresholds, corresponding to a false alarm rate of 1/month.