Rapid and robust parameter inference for binary mergers

The detection rate for compact binary mergers has grown as the sensitivity of global network ground based gravitational wave detectors improved, now reaching stage where robust automation analyses is essential. Automated low-latency algorithms have been developed that send out alerts when candidate signals are detected. include sky maps to facilitate electromagnetic follow-up observations, along with probabilities system might contain a neutron star, and hence be more likely generate an counterpart. Data quality issues, such loud noise transients (glitches), can adversely affect algorithms, causing false alarms throwing off parameter estimation. Here new analysis method presented against glitches, capable producing fully Bayesian inference, including mass estimates, in matter minutes. Key elements wavelet-based de-noising, penalized maximization likelihood during initial search, rapid localization using precomputed inner products, heterodyned likelihoods full inference.