Efficient One-Way Wave-Equation Depth Migration Using Fast Fourier Transform and Complex Padé Approximation via Helmholtz Operator

“One-way wave-equation depth migration (OWDM) is a foundational and efficiency-driven technique in seismic imaging. However, conventional OWDM methods, such as the split-step Fourier (SSF) and generalized screen propagator (GSP) methods, exhibit limited accuracy, particularly when imaging geological structures with steeply dipping reflectors or strong lateral velocity variations. Highly accurate local-velocity-based approaches are prohibitive in 3D applications due to computational intensity for large-scale tensor decomposition operations or multiple iterations. To address this critical issue, this study proposes a novel OWDM method combining the Fast Fourier Transform (FFT) and Padé approximation for wavefield extrapolation (FFT–Padé). The method efficiently calculates the square-root operator in the wavenumber domain using a pseudo-spectral Helmholtz operator, then approximates the one-way wave propagator with an accurate, low-order Padé expansion. Quantitative numerical tests demonstrate that the proposed FFT–Padé method provides significantly improved amplitude fidelity and effective artifact suppression, notably enhancing the imaging of steep-dip reflectors compared with both SSF and GSP techniques. Critically, these accuracy gains are achieved with minimal computational overhead, as only 16% slower than the GSP method. Application to real seismic data confirms the method’s robustness, superior imaging clarity, with a signal-to-noise ratio approximately 12% greater than that of the SSF method.