Joint Learning for Spatial Context-based Seismic Inversion of Multiple Datasets for Improved Generalizability and Robustness

TL;DR

This paper introduces a joint learning approach with a temporal convolutional network and spatial context integration for seismic inversion, enhancing robustness and generalization across multiple datasets, especially with noisy data and limited labels.

Contribution

It proposes a novel joint learning scheme and a temporal convolutional network that models seismic traces with spatial context, improving inversion accuracy and robustness.

Findings

Achieved lowest average MSE of 0.0966 on SEAM dataset.

Demonstrated improved generalization across multiple datasets.

Provided robust impedance estimations in noisy conditions.

Abstract

Seismic inversion plays a very useful role in detailed stratigraphic interpretation of seismic data. Seismic inversion enables estimation of rock properties over the complete seismic section. Traditional and machine learning-based seismic inversion workflows are limited to inverting each seismic trace independently of other traces to estimate impedance profiles, leading to lateral discontinuities in the presence of noise and large geological variations in the seismic data. In addition, machine learning-based approaches suffer the problem of overfitting if there is a small number of wells on which the model is trained. We propose a two-pronged strategy to overcome these problems. We present a Temporal Convolutional Network that models seismic traces temporally. We further inject spatial context for each trace into its estimations of the impedance profile. To counter the problem of limited labeled data, we also present a joint learning scheme whereby multiple datasets are simultaneously used for training, sharing beneficial information among each other. This results in the improvement in generalization performance on all datasets. We present a case study of acoustic impedance inversion using the open source SEAM and Marmousi 2 datasets. Our evaluations show that our proposed approach is able to perform robust and laterally consistent estimations of impedance in the presence of noisy seismic data and limited labels. We compare and contrast our approach to other learning-based seismic inversion methodologies in the literature. On SEAM, we are able to obtain an average MSE of 0.0966, the lowest among all other methodologies.