High Accuracy Fits for Merger Remnants and Time-Dependent Parameter Surrogates for Eccentric Unequal-Mass Non-Spinning Black-Hole Binaries

Over the past decade, gravitational-wave detections from LIGO–Virgo–KAGRA have opened up a new avenue to study black hole astrophysics, offering unprecedented insights into the formation, dynamics, and final states of merging compact objects. For these analyses, models for waveforms and remnant quantities from binary black hole (BBH) coalescences are instrumental. We present a model for accurate fits for the remnant properties NRSurE_q4NoSpin_Remnant, for unequal-mass non-spinning eccentric binary black holes, trained on banks of numerical-relativity simulations. For this we use Gaussian process regression (GPR) to interpolate for the remnant mass, spin, and recoil velocity in the 3-dimensional parameter space with mass ratios q≤4, eccentricity e<0.25, and include the impact of the mean anomaly, a cyclic parameter l ∈ [0, 2π), both defined at particular reference times before merger. Being trained directly on eccentric numerical relativity simulations, our fits are free from ambiguities regarding the initial frequency at which eccentric quantities are defined. As a byproduct of using GPR, we also provide error estimates for all modeled quantities, which can be consistently incorporated into current and future gravitational-wave parameter-estimation analyses. We also present NRSurE_q4NoSpin_Aux, a time-domain surrogate model for eccentricity and mean anomaly evolution, trained on the same bank of numerical relativity simulations, to map the eccentric parameters to different reference times before merger. Together, these models provide a more complete and physically consistent description of eccentric non-spinning black hole mergers, helping improve future analyses of GW signals.