Project Grant 2607932

The National Science Foundation Division of Physics awarded the University of Rhode Island $150,000 on August 1, 2026, under the Mathematical and Physical Sciences program (CFDA 47.049) for a project grant to develop machine learning approaches for exploring binary black hole parameter space and quantifying accuracy limitations in numerical relativity simulations used to predict gravitational waveforms.

The award funds three primary deliverables. First, the recipient will create a neural network to predict waveform similarity across eccentric binary black hole systems, enabling more efficient and comprehensive exploration of this parameter space. Second, the project will quantify systematic biases and accuracy limitations in current numerical relativity simulations by injecting high-resolution waveforms as mock gravitational wave signals and performing parameter estimation using lower-resolution simulation suites. Third, the recipient will expand the MAYA catalog of numerical relativity waveforms to support improved gravitational wave detection and interpretation as detector sensitivity advances.

The work is performed in Kingston, Rhode Island. The period of performance runs from August 1, 2026, through July 31, 2029. These improved waveform predictions and simulation accuracy assessments are intended to enable gravitational-wave observatories to interpret high signal-to-noise ratio observations from merging black holes as detector technology continues to improve in sensitivity and frequency range.