How to Catch Acceleration by the Tail express
An optical trap holds an atom almost motionless, like a ball on a soft spring. But as soon as acceleration appears, say gravity, the equilibrium shifts. The atom moves within the standing light wave, and this changes the rate at which it exchanges energy with the field. The frequency of this exchange becomes a direct indicator of acceleration.
The authors calculated a model where the atom and field start from the most tranquil state, then exactly solved the equations of motion. It turns out that information about acceleration accumulates in the joint state of the atom and light. It can be extracted, and the longer you wait, the more precise it gets — precision grows quadratically with time.
There are different ways to read out the result. At a node of the standing wave, the phase method works: it senses the sign of acceleration. Away from the node, it's better to count photons or use homodyne detection with a tweaked phase. Even if the cavity has losses, there's an optimal measurement time when noise hasn't yet eaten all the useful information. This approach is like weighing an invisible load by the jitter of a spring — except the role of the spring is played by the vacuum structure itself.
Original abstract
as written by the authors arXiv:2607.23324
We derive an effective trapped atom--cavity model in which a constant gravitational acceleration shifts the oscillator equilibrium and changes the local standing-wave coupling, thereby encoding the acceleration in the Jaynes Cummings exchange frequencies. With the atom and motion initially in their ground states and the cavity field initially coherent, we solve the closed-system carrier dynamics exactly and derive the displaced-frame quantum Fisher information (QFI) of the joint atom-cavity state. This QFI is proportional to the square of the local coupling slope, grows quadratically with interrogation time, and scales linearly with mean photon number. At the node, phase-referenced Ramsey detection gives a sign-sensitive estimate of axial acceleration and locally saturates the joint QFI. Away from the node, photon counting and phase-optimized homodyne detection provide cavity readouts when the cavity state carries more QFI than the atomic state. At the off-node operating point studied, Lindblad simulations show that cavity loss produces a finite-time QFI optimum. Lamb Dicke and sideband-suppression conditions control the carrier approximation. In the closed-system benchmark, the carrier-model QFI agrees with the atom-cavity QFI obtained from the unexpanded model after tracing out motion.
The shift in the atom's equilibrium in the trap due to gravity is roughly the diameter of a proton, but that's enough to noticeably change the frequency of light exchange.