Source separation using multistage inversion with sparsity promoting priors

1 . A method, comprising:

acquiring blended seismic data representing a subsurface volume of interest from a plurality of seismic sources;

estimating a first plurality of signal modes using a plurality of priors by applying sparse inversion to the blended seismic data;

selecting a first strongest signal mode from the first plurality of signal modes;

predicting first multi-source interference in the blended seismic data based at least in part on the first strongest signal mode;

separating the first strongest signal mode and the first multi-source interference from the blended seismic data, such that a first residual signal is left;

displaying a first image of the first residual signal;

based on the first image, identifying whether the first residual signal is deblended;

responsive to the first residual signal not being deblended, estimating a second plurality of signal modes using the plurality of priors by applying the sparse inversion to the first residual signal;

selecting a second strongest signal mode from the second plurality of signal modes;

predicting second multi-source interference in the blended seismic data based at least in part on the second strongest signal mode;

separating the second strongest signal mode and the second multi-source interference from the first residual signal, such that a second residual signal is left;

displaying a second image of the second residual signal;

based on the second image, identifying whether the second residual signal is deblended; and

responsive to the second residual signal being deblended, estimating a coherent signal from the second residual signal by solving the sparse inversion.

2 . The method of claim 1 , further comprising generating an image representing the subsurface volume of interest based at least in part on the coherent signal.

3 . The method of claim 1 , wherein estimating the first plurality of signal modes includes using a sparsity inversion promoting transform that is multi-dimensional.

4 . The method of claim 1 , wherein the sparse inversion comprises at least one of exploiting a sparsity or low-rank structure of seismic data.

5 . The method of claim 1 , wherein a first prior of the plurality of priors is configured to increase a first sparsity of a first signal mode of the first plurality of signal modes, and wherein a second prior of the plurality of priors is configured to increase a second sparsity of a second signal mode of the first plurality of signal modes.

6 . The method of claim 5 , wherein the first signal mode comprises a direct arrival, and wherein the second signal mode comprises a reflection, refraction, a coherent noise component, or a combination thereof.

7 . The method of claim 1 , wherein the blended seismic data includes one or both of pressure motion measurements or particle motion measurements.

8 . The method of claim 1 , wherein the plurality of priors includes at least one of noise attenuation, timing information of the blended seismic data, or frequency bands in the blended seismic data.

9 . The method of claim 1 , wherein the plurality of priors includes velocity model data representing propagation characteristics through the subsurface volume.

10 . The method of claim 9 , wherein the plurality of priors includes one or both of a moveout correction or static correction.

11 . The method of claim 1 , further comprising, before identifying whether the first residual signal is deblended, separating a third strongest signal mode and third multi-source interference from the blended seismic data to form the first residual signal.

12 . A computing system, comprising:

one or more processors; and

a memory system comprising one or more non-transitory, computer-readable media storing instructions that, when executed by at least one of the one or more processors, cause the computing system to perform operations, the operations comprising:

acquiring blended seismic data representing a subsurface volume of interest from a plurality of seismic sources;

estimating a first plurality of signal modes using a plurality of priors by applying sparse inversion to the blended seismic data;

selecting a first strongest signal mode from the first plurality of signal modes;

predicting first multi-source interference in the blended seismic data based at least in part on the first strongest signal mode;

separating the first strongest signal mode and the first multi-source interference from the blended seismic data, such that a first residual signal is left;

displaying a first image of the first residual signal;

based on the first image, identifying whether the first residual signal is deblended;

responsive to the first residual signal not being deblended, estimating a second plurality of signal modes using the plurality of priors by applying the sparse inversion to the first residual signal;

selecting a second strongest signal mode from the second plurality of signal modes;

predicting second multi-source interference in the blended seismic data based at least in part on the second strongest signal mode;

separating the second strongest signal mode and the second multi-source interference from the first residual signal, such that a second residual signal is left;

displaying a second image of the second residual signal;

based on the second image, identifying whether the second residual signal is deblended; and

responsive to the second residual signal being deblended, estimating a coherent signal from the second residual signal by solving the sparse inversion.

13 . The computing system of claim 12 , wherein the operations further comprise generating an image representing the subsurface volume of interest based at least in part on the coherent signal.

14 . The computing system of claim 12 , wherein a first prior of the plurality of priors is configured to increase a first sparsity of a first signal mode of the first plurality of signal modes, and wherein a second prior of the plurality of priors is configured to increase a second sparsity of a second signal mode of the first plurality of signal modes.

15 . The computing system of claim 14 , wherein the first signal mode comprises a direct arrival, and wherein the second signal mode comprises a reflection, refraction, a coherent noise component, or a combination thereof.

16 . The computing system of claim 12 , wherein the blended seismic data includes one or both of pressure motion measurements or particle motion measurements.

17 . The computing system of claim 12 , wherein the plurality of priors includes at least one of noise attenuation, timing information of the blended seismic data, or frequency bands in the blended seismic data.

18 . The computing system of claim 17 , wherein the plurality of priors includes one or both of a moveout correction or static correction.

19 . A non-transitory, computer-readable medium storing instructions that, when executed by at least one processor of a computing system, cause the computing system to perform operations, the operations comprising:

acquiring blended seismic data representing a subsurface volume of interest from a plurality of seismic sources;

estimating a first plurality of signal modes using a plurality of priors by applying sparse inversion to the blended seismic data;

selecting a first strongest signal mode from the first plurality of signal modes;

predicting first multi-source interference in the blended seismic data based at least in part on the first strongest signal mode;

separating the first strongest signal mode and the first multi-source interference from the blended seismic data, such that a first residual signal is left;

displaying a first image of the first residual signal;

based on the first image, identifying whether the first residual signal is deblended;

responsive to the first residual signal not being deblended, estimating a second plurality of signal modes using the plurality of priors by applying the sparse inversion to the first residual signal;

selecting a second strongest signal mode from the second plurality of signal modes;

predicting second multi-source interference in the blended seismic data based at least in part on the second strongest signal mode;

separating the second strongest signal mode and the second multi-source interference from the first residual signal, such that a second residual signal is left;

displaying a second image of the second residual signal;

based on the second image, identifying whether the second residual signal is deblended; and

responsive to the second residual signal being deblended, estimating a coherent signal from the second residual signal by solving the sparse inversion.