Preparing for the fourth observing run of advanced LIGO-Virgo: Search for binary black hole mergers with the updated Coherent WaveBurst pipeline*

The advanced LIGO-Virgo detectors have discovered gravitational wave (GW) signals from over 90 binary black hole mergers (BBH). With improved sensitivities of the advanced LIGO-Virgo detectors in the upcoming LIGO-Virgo-KAGRA (LVK) fourth observing run (O4), many more GW events are expected to be detected. In particular, we are interested in detecting exceptional high-mass BBH mergers, similar to the GW190521 event discovered in 2021 that provided the first direct evidence of an intermediate-mass black hole (IMBH). Since the signals from unprecedented events may not be well-modeled, template-independent search methods like the Coherent WaveBurst (cWB) are used in LVK data analysis. cWB can identify generic GW signals in the LIGO-Virgo data with minimal assumptions on the signal model. The cWB algorithm looks for excess power events in the time-frequency domain and uses a machine learning (ML) method to enhance the signal-noise classification of the identified events. For the O4 run, the cWB search was upgraded with a new time-frequency transform called wavescan that provides high-resolution time-frequency distributions for efficient recovery of transient GW signals. The upgraded cWB search allows for precise estimation of the signal properties, consequently improving the sensitivity of the ML-enhanced cWB search to BBH mergers. This study discusses the results of the proposed O4 cWB search on LIGO-Virgo data from the third LVK observing run (O3). We report GW detections with higher significance and demonstrate improvement in the detection efficiency by approximately 40% for simulated BBH events at a false alarm rate of less than one per year.