Abstract:SUMMARY The receiver function (RF) method is one of the most widely used passive seismic approaches for imaging subsurface structures from shallow sedimentary basins to the deep Earth’s interior. RF is typically computed with a deconvolution operation that enables separating structural response from source–time function embedded in teleseismic wavefields. However, seismic recordings of temporary arrays are often contaminated with strong random noise and even amplitude bias due to poor installation conditions. Both factors can cause instability to deconvolution and severely degrade the accuracy, thereby the imaging quality, of the RF method. This underscores the importance of effective denoising algorithms in RF processing. In this study, we explore the application of high-resolution Radon transform (RT) to improve the conventional RF imaging workflow. Contrary to the commonly implemented post-processing (i.e. after deconvolution) noise suppression schemes, we introduce the RT to data pre-processing (i.e. before deconvolution). This method seeks a sparse representation of teleseismic wavefields in the τ–p domain by iteratively solving a least-squares minimization problem with the conjugate gradient algorithm. Synthetic test with a 2-D step-Moho model shows that non-linear phase arrivals including incoherent noise and diffraction energy are effectively removed in resulting RFs, with the signal-to-noise ratio increased by as much as ∼8 dB. Real data experiments using the Hi-CLIMB network in the Tibetan Plateau demonstrate the superior performance of the proposed workflow in regularizing the wavefield and improving the coherence of converted phases across the recording array. Consequently, imaging results of common conversion point stacking using a single and a group of teleseismic events both recover subtle converted phases from the Moho and potential lithospheric discontinuities that are otherwise obscured by noise arising from the conventional processing workflow. This study highlights the necessity of wavefield regularization in the RF method and calls for improved data processing techniques in array-based seismic imaging.

Study on parameter choice methods for the RFMP with respect to downward continuation

Riemannian Optimization for Phase Retrieval from Masked Fourier Measurements

High-dimensional experiments for the downward continuation using the LRFMP algorithm

Parameter Selection by GCV and a $χ^2$ test within Iterative Methods for $\ell_1$-regularized Inverse Problems

Parametric Refinement of the Rational Function Model for Satellite Remote Sensing Imagery

Optimization of Random Feature Method in the High-Precision Regime

Performance Analysis on Parameter Estimation of Orthogonal Matching Pursuit Based on the Attribute Scattering Center Model

On Learning the Optimal Regularization Parameter in Inverse Problems

A discretizing Tikhonov regularization method via modified parameter choice rules

Choosing Appropriate Regularization Parameters by Splitting Data into Training and Validation Sets—Application in Global Surface-Wave Tomography

Off-the-grid regularisation for Poisson inverse problems

An Adaptive Iterative Method for Downward Continuation of Potential-Field Data from a Horizontal Plane

General Optimization Framework for Robust and Regularized 3D Full Waveform Inversion

Parameter Inversion Method of Multilayered Media Based on Off-Grid Sparse CMP Model With Refined Orthogonal Matching Pursuit

Adaptive Reduced Basis Trust Region Methods for Parameter Identification Problems

Adaptive selection strategy of shape parameters for LRBF for solving partial differential equations

Adaptive greedy algorithms based on parameter-domain decomposition and reconstruction for the reduced basis method

A modified PRP conjugate gradient method with inertial extrapolation for sparse signal reconstruction

Application of the $χ^2$ principle and unbiased predictive risk estimator for determining the regularization parameter in 3D focusing gravity inversion

Improving Receiver Function Imaging with High-Resolution Radon Transform

Multi-Target Parameter Estimation of the FMCW-MIMO Radar Based on the Pseudo-Noise Resampling Method