Prospects for High-Frequency Gravitational-Wave Detection with GEO600

TL;DR

The paper addresses the challenge of detecting high-frequency gravitational waves by proposing detuning GEO600's signal-recycling mirror to create tunable, narrow-band resonances that can scan into the tens of kHz range. It evaluates GEO600’s high-frequency sensitivity with detuned configurations, comparing against LIGO and Cosmic Explorer, and analyzes two primary source classes: ultralight boson clouds around spinning black holes and sub-solar-mass mergers. The results show GEO600 can outperform the aLIGO design sensitivity in certain high-frequency bands for vector boson-cloud signals (up to about 31 kHz with SNR≈8 in a week) but not for scalar bosons or most SSM mergers, making GEO600 a promising near-term probe of new physics in the Milky Way. The work highlights practical paths for high-frequency GW astronomy and motivates further development in detuning control, time-domain simulations, and squeezing techniques.

Abstract

Current ground-based interferometers are optimized for sensitivity from a few tens of Hz to about 1 kHz. While they are not currently utilized for GW detection, interferometric detectors also feature narrow bands of strong sensitivity at higher frequencies where the sideband fields created by a GW are resonantly amplified in the optical system. For certain interferometer configurations, small changes to system parameters allow the narrow band of high sensitivity to be scanned over a much larger range of frequencies, potentially enabling broadband detection at high frequencies. In this paper, we investigate whether simply modifying the detuning angle of the signal-recycling mirror of the GEO600 interferometer can make this experiment sensitive to GWs in the kilohertz frequency range. We compute the strain sensitivity for GEO600 across a frequency range from several kHz to tens of kHz for various detuning angles. We also show that LIGO cannot attain the same effect assuming that the optical components are not changed due to the narrow band response of the Fabry-Perot cavities. We then calculate the sensitivity of GEO600 to various proposed high-frequency GW sources and compare it to the sensitivity of other ground-based detectors.

Figures (8)

• Figure 1: GEO600 optical layout used in Finesse model. FI: Faraday isolator, MPR: power-recycling mirror, MSR: signal-recycling mirror, BS: beam splitter, PD: photodiode for DC readout. MFE and MCE are the far east mirror and central east mirror. $\phi$ is the detuning angle of the MSR in degrees of a wavelength. The notation for the north arm is analogous.
• Figure 2: Strain sensitivity $A_h(f)$ of GEO600 at different detuning angles of the MSR, as well as a scanned sensitivity curve (black), which is a summation of all possible MSR detuning angles. As the detuning angle of the MSR is increased from $\phi=0$ (tuned configuration for GEO600), the detection region becomes increasingly narrow-band, in that the frequency bandwidth stays constant while the SRC resonant frequency increases. The scanned sensitivity curve is plotted to show the possible values that can be achieved across the kHz frequency range by shifting the MSR.
• Figure 3: Strain sensitivity $A_h(f)$ of the aLIGO design at different detuning angles of the MSR. When the angle of the MSR is decreased from $\phi=90$ (tuned configuration for aLIGO that corresponds to 'signal extraction' Buonanno:2001cjHeinzel:1999rk), a region of strong sensitivity gradually shifts to lower frequencies in the neighborhood of $f\sim \mathcal{O}(100~\text{Hz})$, while its bandwidth simultaneously narrows. Note that we use $\phi=1$ to represent the 'signal recycling' limit, as setting $\phi=0$ exactly corresponds to a configuration completely dominated by classical noise. There is no enhancement in sensitivity at frequencies in the kHz from aLIGO detuning the MSR. The peaks in sensitivity occurring at integer multiples of $37.5$ kHz (FSR) are created by the Fabry-Perot cavities and are independent of the MSR detuning angle. Note that in this plot we are using current LIGO mirrors --- it is possible that changing the MSR or TM transmissivities then LIGO could have improved high-frequency sensitivity.
• Figure 4: Strain sensitivity $A_h(f)$ of GEO600 tuned (purple), anti-tuned (gray), and scanned (dashed), which is a summation of all possible MSR detuning angles. The aLIGO design sensitivity (red) 2020ascl.soft07020R and Cosmic Explorer's design sensitivity (light blue) 2020ascl.soft07020R are also shown. GEO600's scanned sensitivity curve shows the possible frequencies in the kHz range where it would have an advantageous narrow band detection to that of Cosmic Explorer and aLIGO.
• Figure 5: Left: contour lines corresponding to a SNR of eight for GWs sourced from both vector (solid) and scalar (dashed) boson clouds. The contour lines correspond to the strain sensitivities of aLIGO (blue), tuned GEO600 (orange), and GEO600 with different MSR detuning angles. The shaded gray region corresponds to the parameter space where the superradiance condition fails. An SNR of eight can be achieved at frequencies as large as $31$ kHz, with the highest frequency obtained at a detuning angle of $\phi = 45\degree$ (not plotted). Right: SNR of GWs sourced from both vector (solid) and scalar (dashed) boson clouds across a range of boson rest energies for a 0.3 $M_\odot$ BH. The SNR curves are calculated with the same strain sensitivities as the left plot and have the same corresponding colors. The dotted black line represents where the SNR is eight. The sharp increase in SNR in aLIGO at 37.5 kHz (FSR) arises due to sideband resonance in its Fabry-Perot cavities, while the kink in the SNR function on the right plot around 14 kHz corresponds to the sudden drop-off in integration time when $\tau_\text{GW}$ falls below one week. All GWs are sourced from within the galaxy, being simulated from a distance of $30$ kPc.