Trace-wise coherent noise suppression via a self-supervised blind-trace deep-learning scheme

Abstract

Seismic data denoising via supervised deep learning is effective and popular but requires noise-free labels, which are rarely available. Blind-spot networks circumvent this requirement by training directly on noisy data and have been shown to be a powerful suppressor of random noise. In this work, we expand the methodology of blind-spot networks to create a blind-trace network that successfully removes trace-wise coherent noise. An extensive synthetic analysis illustrates the denoising procedure’s robustness to varying noise levels and varying numbers of noisy traces within shot gathers. It is shown that the network can accurately learn to suppress the noise when up to 60% of the original traces are noisy. Furthermore, the proposed procedure is implemented on the Stratton 3D field dataset and is shown to restore the previously corrupted direct arrivals. In addition to trace-wise noise suppression, we adapted the blind-spot networks to the successful suppression of colored Gaussian noise, which exhibits varying coherent properties in both time and spatial axes. The proposed adaptation of the blind-spot networks paves the way for its use in other applications, such as the suppression of coherent noise arising from wellsite activity, passing vessels, or nearby industrial activity.

Semi‐blind‐trace algorithm for self‐supervised attenuation of trace‐wise coherent noise

“…A similar performance can be seen in figure 3 from S. Liu et al. (2022). We alleviate these issues by the method proposed in the following section.…”

“…The process of replacing parts of the data with noise before feeding the data to a DNN is called ‘masking’. In its seismic adaption (Wang, Hu, et al., 2022; S. Liu et al., 2022; Abedi et al., 2023), randomly selected traces in the input data are masked and then a DNN is trained to reconstruct them. Since the network is blinded to a trace in the data, the method is called ‘blind‐trace denoising’.…”

“…An extension of this method (named structured noise to void) aims to attenuate noise along extended blind masks (Broaddus et al., 2020), in which the noise is deemed coherent. In seismic studies, adaptations of these methods are used for the attenuation of random noise (Meng et al., 2021; Birnie et al., 2021), trace‐wise coherent noise (Birnie & Alkhalifah, 2022; S. Liu et al., 2022; Abedi et al., 2023) and deblending (Wang, Hu, et al., 2022).…”

“…A key requirement for denoising methods is to preserve the signal quality while addressing the denoising task (Ovcharenko and Hou, 2020). In blind‐trace denoising (S. Liu et al., 2022), all traces – whether noisy or not – are reconstructed from the neighbouring traces. This results in signal leakage in the entire data and reconstruction errors close to the noisy traces.…”

Explainable artificial intelligence‐driven mask design for self‐supervised seismic denoising

“…In seismic applications, Liu et al. (2022) used prior knowledge of the noise's spatio‐temporal characteristics to design a trace‐wise noise mask. Similarly, Wang et al.…”