Can Gravito-Magnetism Separate Low-Frequency Gravitational Waves from Asteroid Noise?

Asteroid Gravity Gradient Noise Dominates the Low-Frequency Band

In the low-frequency band from 10 nHz to 1 micro-Hz, the paper states that any detector using proof masses inside the inner solar system faces a specific obstacle: the time-varying gravitational field produced by the random motions of asteroids in the asteroid belt. This is called gravity gradient noise, or GGN. The paper puts its size at 2 to 4 orders of magnitude larger than the target gravitational wave signal.

That ratio explains why the problem is not a matter of building a quieter instrument. The noise is gravitational, so it moves a freely falling proof mass in the same way a signal would. The paper notes that pulsar timing arrays avoid this noise because their proof masses, the pulsars, sit far outside the solar system, and that proposed white-dwarf astrometry would also rely on distant proof masses. Both routes carry their own instrumental burdens, which is the motivation for finding a way to keep the proof masses closer.

Relativistic Waves Carry Equal Electric and Magnetic Parts

The paper's central idea comes from an analogy with electromagnetism. A relativistic electromagnetic wave has electric and magnetic field amplitudes of equal size. A stray, non-relativistic moving charge instead produces a magnetic field suppressed by its velocity. A sensor that responds to magnetic fields can therefore pick out the wave from the local charge background.

The authors apply the same logic to gravity. A relativistic gravitational wave has gravito-electric and gravito-magnetic components of equal amplitude. The time-varying gravitational field of moving asteroids is dominantly gravito-electric, and its gravito-magnetic part is suppressed by the asteroid velocity. A detector sensitive to the gravito-magnetic effect can therefore separate a genuine wave from the local GGN background. The paper is explicit that this is a proposed detection concept, not a demonstration with data.

The Asteroid Velocity Sets the Suppression Ceiling

How much separation is available depends on how fast the asteroids move. The paper uses the orbital velocity of the asteroid belt, about 10^-4 in units of the speed of light, as a conservative estimate of the maximum signal-to-noise suppression achievable. It notes that the asteroids share this bulk motion around the Sun, and that the variance within that bulk flow is the actual source of the GGN.

That number matters because the expected gravitational wave signals are only 2 to 4 orders of magnitude weaker than the asteroid GGN. A suppression factor set by 10^-4 is therefore described as more than sufficient to permit signal extraction. The comparison is between the asteroid-induced gravito-magnetic component and the wave's gravito-magnetic component, not between the total GGN and the total wave amplitude.

Lorentz Acceleration Makes the Suppressed Signal Measurable

The paper considers two ways to sense gravito-magnetism. The first is spin precession, where the smallest detectable characteristic strain scales with the inverse square of frequency and interrogation time and with the inverse square root of the number of spins. The authors state they could not find an experimentally realizable spin configuration that would detect the expected precession in this band.

They instead focus on the velocity-dependent Lorentz force acceleration on test bodies. Here the shot-noise-limited strain scales with the ratio of probe wavelength to baseline, divided by frequency squared and interrogation time squared, and by the inverse square root of the number of photons or atoms. The paper argues that the wavelength-to-baseline ratio provides the metrological boost that makes the Lorentz force sensitive enough. Maximizing the signal requires interrogation times of order the inverse gravitational wave frequency, so the proof masses separate by roughly the relative velocity times that time.

The paper also states why this pushes the design into space. A large baseline and relative velocity between test bodies boost the signal, and space can accommodate the large separation that long interrogation times imply. On Earth, free fall with atoms lasts at most about 10 seconds, far shorter than the period of the waves of interest.

Three Satellites Cancel Higher-Order Gravity Gradient Noise

The proposed configuration uses three satellites in elliptical orbits around the Sun, kept within about 10 million km of each other. One satellite serves as the master clock, and the local clocks of the others are synchronized to it at an initial time. The satellites exchange laser pulses at designated global times, and the receiving satellite computes a discrete double time difference from arrival times at three successive moments.

That double difference isolates gravitational signatures from initial conditions such as position and velocity. A single such term contains both the gravity gradient and the gravitational wave signal, so the paper uses additional baselines to invert the signal matrix and cancel the higher-order GGN. The authors report that this full reconstruction with algebraic cancellation is possible with three satellites without severe signal-to-noise penalties.

The paper is careful about what this does and does not establish. It presents the measurement protocol as the primary result, and states that gravito-magnetism does not by itself solve other metrological problems, such as building proof masses sensitive only to gravity or achieving the required rotational stability.

Pointing Error Changes the Satellite Count

Pointing error is a separate noise source: if a satellite sends light slightly off the intended direction, the arrival-time shift can mimic an acceleration. The paper notes that this error originates at the emitting satellite, so every baseline sharing that emitter inherits the same error, which makes it distinguishable from the gravito-magnetic acceleration.

In simulation, including this noise increased the signal-to-noise penalty by a factor of about 100 for a three-satellite baseline. In a four-satellite constellation the penalty dropped to order one. The authors therefore conclude that while cancelling gravity gradient noise needs no more than three satellites, relaxing the pointing requirement makes a fourth satellite likely necessary. As an alternative, they suggest a guide star to define the pointing axis, which would require pointing accuracy of about 10^-11 radians, roughly a microarcsecond, comparable to the astrometric precision of missions such as GAIA.

Atom Interferometers Are Proposed as the Clean Proof Mass

Even with the noise-cancellation scheme, the measurement needs a proof mass that responds exclusively to gravity at ultra-low frequencies. The paper proposes ultra-cold atom interferometers as a pristine inertial reference and treats a full analysis of such systems as future work.

The authors examine several noise channels for this proof mass. They describe a trade-off in which changing the interferometer operation distance to about 100 m suppresses one noise term by a factor of roughly 27, with further suppression possible from more atoms, more cooling, or a slightly shorter interrogation time. They also suggest that a more elaborate pulse sequence could cancel phases that depend linearly on thermal velocity while retaining sensitivity to the satellite's acceleration. Position fluctuations of the atom cloud are stated to stay below shot noise as long as they are controlled to within about a millimeter.

The paper also estimates local stability requirements, assuming on-board telescopes provide local pointing accuracy of about 10^-13 per second averaged over a measurement bandwidth of about 10^-7 Hz, which it describes as a factor of 1000 worse than the stability demonstrated by Gravity Probe B. At that level, residual centrifugal accelerations out to about 100 m are smaller than the required sensitivity, and the Coriolis acceleration on a pico-kelvin cloud of about 10^10 atoms is comparable to the required shot-noise sensitivity.

The Same Architecture Is Easier at Microhertz Frequencies

The paper closes by noting that the hybrid architecture, a local atom interferometer continuously stabilizing a macroscopic LISA-like proof mass, can also be deployed in the micro-Hz to milli-Hz band. There, asteroid GGN is negligible, so detectors can target the unsuppressed gravito-electric signal and the metrological demands on the atom interferometer relax.

The paper gives a concrete comparison. Reaching an expected strain of about 10^-18 at 1 micro-Hz requires an atom interferometer to operate for only about 10^3 seconds over a spatial drift of about 3 m, with a phase sensitivity of 10^-3 per root Hz, assuming a baseline of about 10^6 km. The authors describe these parameters as remarkably less stringent than those required for the nHz gravito-magnetic detector, and also weaker than the atom-sensing capabilities demanded by mid-band detectors near 1 Hz, where intrinsic source strains are much lower.

The paper also points to ultra-cold molecules and nanoparticle interferometry as ways to shrink the apparatus. It estimates that laser-cooled molecules with a mass around 1000 GeV would need to traverse only meters rather than the roughly 30 m the atoms travel, allowing a much more compact setup. These remain directions for future work rather than demonstrated designs.

The priority the paper establishes is therefore a split by frequency. In the nHz band, the gravito-magnetic channel is what makes inner-solar-system proof masses viable against asteroid GGN, at the cost of a velocity-suppressed signal and demanding atom-interferometer stability. At micro-Hz frequencies, the same hardware can pursue the stronger gravito-electric signal with far looser requirements, which the authors present as the more immediately promising application.

Sources used in this answer