Implementing a search for gravitational waves from binary black holes with nonprecessing spin

Abstract

Searching for gravitational waves (GWs) from binary black holes (BBHs) with LIGO and Virgo involves matched-filtering data against a set of representative signal waveforms -a template bankchosen to cover the full signal space of interest with as few template waveforms as possible. Although the component black holes may have significant angular momenta (spin), previous searches for BBHs have filtered LIGO and Virgo data using only waveforms where both component spins are zero. This leads to a loss of signal-to-noise ratio for signals where this is not the case. Combining the best available template placement techniques and waveform models, we construct a template bank of GW signals from BBHs with component spins χ1,2 ∈ [−0.99, 0.99] aligned with the orbital angular momentum, component masses m1,2 ∈ [2, 48] M , and total mass M total ≤ 50 M . Using effectiveone-body waveforms with spin effects, we show that less than 3% of the maximum signal-to-noise ratio (SNR) of these signals is lost due to the discreetness of the bank, using the early advanced LIGO noise curve. We use simulated advanced LIGO noise to compare the sensitivity of this bank to a non-spinning bank covering the same parameter space. In doing so, we consider the competing effects between improved SNR and signal-based vetoes, and the increase in the rate of false alarms of the aligned-spin bank due to covering a larger parameter space. We find that the aligned-spin bank can be a factor of 1.3 -5 more sensitive than a non-spinning bank to BBHs with dimensionless spins > +0.6 and component masses 20 M . Even larger gains are obtained for systems with equally high spins but smaller component masses.

Binary Black Hole Mergers in the First Advanced LIGO Observing Run

“…The search was performed using two independently implemented analyses, referred to as PyCBC [2][3][4] and GstLAL [5][6][7]. These analyses use a common set of template waveforms [8][9][10] but differ in their implementations of matched filtering [11,12], their use of detector data-quality information [13], the techniques used to mitigate the effect of non-Gaussian noise transients in the detector [5,14], and the methods for estimating the noise background of the search [3,15]. We obtain results that are consistent between the two analyses.…”

Upper Limits on the Rates of Binary Neutron Star and Neutron Star–black Hole Mergers From Advanced Ligo’s First Observing Run

“…The offline search in O1 forms a single search targeting BNS, NSBH, and BBH systems. The waveform filters cover systems with individual component masses ranging from 1 to 99 M e , total mass constrained to less than 100 M e (see Figure 1), and component dimensionless spins up to±0.05 for components with mass less than 2 M e and ±0.99 otherwise (Abbott et al 2016c;Capano et al 2016). Waveform filters with total mass less than 4 M e (chirp mass less than 1.73 M e 144 ) for PyCBC (GstLAL) are modeled with the inspiral-only, post-Newtonian, frequency-domain approximant "TaylorF2" (Arun et al 2009;Bohé et al 2013Bohé et al , 2015Blanchet 2014;Mishra et al 2016).…”

Optimizing spinning time-domain gravitational waveforms for advanced LIGO data analysis

“…Our optimization strategies have been shown to dramatically reduce the SEOBNRv2 run-time and over-sensitivity to initial conditions, while maintaining amplitude-weighted phase agreement with the original code to within 0.00790 rad (Table 3), which is entirely dominated by roundoff errors. The work discussed in this paper is hoped to not only assist MCMC parameter estimation but be useful to others in the community who cannot use pre-optimized SEOBNR codes due to their slow speeds (e.g., the generation of stochastic template banks [33]). Some of our optimizations have already been applied to the SEOBNRv3 code within LALSuite, leading to speed-ups of around 15x over its original version.…”