Quantum-optimized phase closure for robust interferometric imaging

Abstract

Closure phase enables imaging with ground-based interferometric arrays by canceling telescope-dependent phase errors due to atmospheric fluctuations. It can be obtained by measuring fringe phases on individual baselines forming a triangle and summing them. We show that closure phases can be measured more accurately with fewer photons by performing joint measurements on multiple photons. We first establish that at least three photons from the same atmospheric temporal phase-stability window are required for closure-phase information to survive the unknown telescope-dependent phase noise, and show that joint three-photon measurements already provide an advantage over single-photon measurements in the minimal three-telescope setting. We then turn to large arrays and consider estimating all bispectra---complex quantities whose phases are the closure phases---under a global mean-square error criterion. Although an $m$-telescope array contains $Θ(m^2)$ independent triangle bispectra, we show that physical constraints on the visibilities make the total information content grow only nearly linearly with $m$ at fixed accuracy. Exploiting this structure, we construct an efficient protocol using at most three-photon joint measurements with photon cost $\tilde{O}(m)$, whereas any sequential single-photon protocol requires $Ω(m^2)$ photons. When the number of detected photons per temporal stability window scales linearly with the array size, this allows the required exposure time to remain independent of $m$ up to logarithmic factors. We further analyze the complementary high-precision regime using multiparameter quantum metrology and find a similar advantage for collective measurements. Overall, our results show that quantum information processing techniques can substantially improve the sensitivity and robustness of optical interferometry.