Turbulence-resilient macroscopic Fourier ptychography

Fourier ptychography (FP) enables synthetic aperture imaging beyond the coherent diffraction limit of the imaging system, yet its extension to long-range scenarios remains fundamentally hindered by atmospheric turbulence. Turbulence introduces stochastic, time-varying wavefront distortions that violate both the deterministic pupil model and the coherent spectral overlapping required by FP, rendering conventional reconstruction strategies physically inconsistent. Here, we propose turbulence-resilient macroscopic Fourier ptychography (TRMFP), a unified computational framework that reformulates FP as a statistically consistent imaging process through random media. TRMFP integrates a speckle-interferometry-inspired preprocessing stage with a closed-loop ptychographic reconstruction. By exploiting ensemble statistics from multiple short-exposure measurements at each aperture position, TRMFP recovers turbulence-suppressed, diffraction-limited sub-aperture images, followed by joint optimization of the object and spatially varying wavefront aberrations under low-rank and total-variation constraints. Simulations and experiments validate the robustness of TRMFP across a wide range of turbulence strengths ( D / r 0 =1.5 to 7.5), achieving ∼2× resolution improvement in 1.5 m phase-screen experiments and ∼1.6× gain under 5 m dynamic turbulence over conventional FP. Without requiring additional hardware, wavefront sensors, or training data, TRMFP establishes a physically grounded and computationally efficient solution for super-resolution imaging through atmospheric turbulence, bridging the gap between Fourier ptychography and real-world long-range synthetic aperture imaging applications.