High-Frequency Thermal Noise in Michelson Interferometers

High-Frequency Thermal Noise in Michelson Interferometers
Daniel Grass, Sander M. Vermeulen, Ian A. O. MacMillan, Lee McCuller
arXiv:2605.16710v2 Announce Type: cross
Abstract: Thermal noise in optics constrains the precision of optical experiments, including Michelson interferometers, optical clocks, and optomechanical sensors. While high-power optical experiments such as gravitational-wave detectors are currently limited mainly by quantum noise, new experiments incorporating recent insights from quantum metrology are being developed that evade the quantum shot noise background. In particular, Michelson interferometers that use photon-counting readout will look for weak, high-frequency signals. Since shot noise is no longer the dominant noise source with these readout schemes, it is important to accurately model thermal noise to characterize signals and design more sensitive experiments. However, previous modeling uses approximations that are no longer valid in these frequency regimes. In the MHz band, the quasistatic approximation does not apply. We therefore develop more general models of substrate and coating mechanical (Brownian) noise, substrate and coating thermoelastic noise, and coating thermorefractive noise. We validate the models with comparisons to previous low-frequency modeling and high-frequency spectra from an experiment that has already taken data, the Holometer. We then apply the new models to GQuEST, an experiment under construction.arXiv:2605.16710v2 Announce Type: cross
Abstract: Thermal noise in optics constrains the precision of optical experiments, including Michelson interferometers, optical clocks, and optomechanical sensors. While high-power optical experiments such as gravitational-wave detectors are currently limited mainly by quantum noise, new experiments incorporating recent insights from quantum metrology are being developed that evade the quantum shot noise background. In particular, Michelson interferometers that use photon-counting readout will look for weak, high-frequency signals. Since shot noise is no longer the dominant noise source with these readout schemes, it is important to accurately model thermal noise to characterize signals and design more sensitive experiments. However, previous modeling uses approximations that are no longer valid in these frequency regimes. In the MHz band, the quasistatic approximation does not apply. We therefore develop more general models of substrate and coating mechanical (Brownian) noise, substrate and coating thermoelastic noise, and coating thermorefractive noise. We validate the models with comparisons to previous low-frequency modeling and high-frequency spectra from an experiment that has already taken data, the Holometer. We then apply the new models to GQuEST, an experiment under construction.