Full text loading...
-
Microseismic Data Reconstruction and Location Based on Bayesian Non-Parametric Dictionary Learning
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 81st EAGE Conference and Exhibition 2019, Jun 2019, Volume 2019, p.1 - 5
Abstract
Microseismic data reconstruction is a procedure to compensate for acquisition deficiencies, and to improve data quality, which is very important for subsequent processing steps, such as event location. Some reconstruction algorithms depend on some parameter settings and ignore the strong noise interference, which may not work well for low quality surface microseismic data. In this paper, we propose to use a Bayesian non-parametric dictionary learning method to recover microseismic signal from the noisy data with missing traces. In the proposed method, the beta-Bernoulli process is employed as a prior for learning an appropriate dictionary to sparsely represent microseismic signals. An approximation to the full posterior is manifested via Gibbs sampling, yielding an ensemble of dictionary and sparse coefficients. Finally, the signal of interest is reconstructed by the product of dictionary and sparse coefficients. Tests on synthetic and real microseismic data demonstrate that the proposed method works very well for low signal-to-noise ratio data with missing traces. We also show how the proposed method benefits reverse time migration based event location.