Extending the Width of Short Aperture Data by Data-Driven Extrapolation

TL;DR

This paper demonstrates that the effective aperture for estimating wave phase velocity can be extended beyond the traditional limit using data-driven extrapolation, improving velocity estimates for surface and body waves.

Contribution

It introduces a data-driven method to extend the recording aperture, enabling more accurate phase velocity estimation from shorter recordings.

Findings

Aperture extension improves velocity estimation accuracy.

Method works for both surface and body waves.

Theoretical and computational validation of the approach.

Abstract

It is commonly believed that it is infeasible to estimate the phase velocity of a surface wave with a recording aperture less than $\lambda/2$ in length, where $\lambda$ is the horizontal wavelength of the propagating wave. We show both theoretically and computationally that the width of the recording aperture can be extended using the recorded data at longer time intervals. This leads to more reliable estimates of apparent velocities for both body waves and surface waves. Here, we assume single-mode plane waves sweeping across the recording array with a linear moveout.