Beyond spatial averaging: Modeling DAS seismic responses with optical system effects

Distributed acoustic sensing (DAS) utilizes standard optical fibers as continuous distributed seismic sensors, enabling the acquisition of subsurface vibration information with high spatial sampling density and long-distance coverage. Existing numerical simulation methods for DAS data typically rely on solving elastic wave equations and are implemented through simplified processes such as axial projection along the fiber and spatial averaging. However, this approach neglects the influence of optical systems on DAS responses, including Rayleigh scattering, coherent detection, and signal demodulation, making it difficult to comprehensively reflect the response characteristics of actual DAS systems. To address this deficiency, this study proposes a DAS seismic response numerical simulation framework that integrates physical processes. We introduce Rayleigh scattering modeling and demodulation mechanisms for coherent OTDR (optical time domain reflectometry) in the simulation, and systematically consider key parameters including pulse width, gauge length, polarization fading, laser frequency drift, phase noise, and detector noise. Simulation results demonstrate that, under ideal noise-free conditions, the new model yields result highly consistent with those of the traditional spatial averaging model, validating its accuracy. In the presence of environmental noise, the proposed method produces smoother responses compared to traditional approaches, indicating a certain denoising advantage. When optical system noise is introduced, the physics-based simulation can reproduce complex phenomena observed in actual measurements, such as phase fading noise and frequency drift noise. These findings underscore the necessity and effectiveness of a physical model in capturing the true behavior of DAS systems. The proposed simulation framework provides a critical foundation for optimizing DAS system design, enhancing signal-to-noise ratios, and improving DAS data inversion. It holds significant guiding implications for the future application of DAS in seismic monitoring.