The eccentricity of binary black hole mergers is predicted to be an indicator of the history of their formation. In particular, eccentricity is a strong signature of dynamical formation rather than formation by stellar evolution in isolated stellar systems. It has been shown that searches for eccentric signals with quasi-circular templates can lead to loss of SNR, and some signals could be missed by such a pipeline. We investigate the efficacy of the existing quasi-circular parameter estimation pipelines to determine the source parameters of such eccentric systems. We create a set of simulated signals with eccentricity up to 0.3 and find that as the eccentricity increases, the recovered mass parameters are consistent with those of a binary with up to a $approx 10%$ higher chirp mass and mass ratio closer to unity. We also employ a full inspiral-merger-ringdown waveform model to perform parameter estimation on two gravitational wave events, GW151226 and GW170608, to investigate this bias on real data. We find that the correlation between the masses and eccentricity persists in real data, but that there is also a correlation between the measured eccentricity and effective spin. In particular, using a non-spinning prior results in a spurious eccentricity measurement for GW151226. Performing parameter estimation with an aligned spin, eccentric model, we constrain the eccentricities of GW151226 and GW170608 to be $<0.15$ and $<0.12$ respectively.
Download