Systems and methods for scaling and thresholding parametrization of shear noise attenuation

A system for removing noise component data comprised in seismic data, the system comprising:

• a computer processor;
• and
• a memory storing a data processing engine that includes instructions which are executable by the computer processor to:
• receive captured seismic data from one or more sensors deployed about a resource site, the captured seismic data including signal component data and the noise component data;
• extract, in a first data domain, velocity data and pressure data included in the captured seismic data and thereby generate first domain data including first pressure data and first velocity data in the first data domain, wherein the first data domain includes a 3-dimensional transform or a 5-dimensional transform, and wherein the first data domain includes a temporal spatial domain;
• transform, from the first data domain to a second data domain, the first velocity data and the first pressure data in the first data domain and thereby generate second domain data including second pressure data and second velocity data in the second data domain, wherein the second data domain includes a Fourier wavenumber domain, wherein transforming, from the first data domain to the second data domain, the first velocity data and the first pressure data comprises performing a Fourier transform along a spatial direction of the first data domain to generate common wavenumber gathers;
• map wavenumber gather data associated with the second pressure data and the second velocity data from the second data domain to a third data domain, wherein the third data domain includes a Fourier-Curvelet domain, and wherein each common wavenumber gather of the Fourier wavenumber domain is mapped to a 2D Curvelet domain;
• apply an enveloping function to the second pressure data and the second velocity data in the third data domain and thereby determine envelope ratio scaling data for the captured seismic data;
• generate, based on noise variance of the second pressure data and the second velocity data, threshold value data for the captured seismic data;
• generate, based on the threshold value data, a thresholding operator for the captured seismic data, wherein the thresholding operator is a stein thresholding operator, a soft thresholding operator, or a combination thereof, and the thresholding operator is applied to the envelope ratio scaling data to generate threshold envelope data;
• apply adaptive localized filtering to the threshold envelope data to generate filter coefficient data from the threshold envelope data for the captured seismic data, wherein the adaptive localized filtering comprises a root-mean-square technique in the third data domain;
• estimate, based on the coefficient data from the threshold envelope data and a weighted linear combination of the second pressure data and the second velocity data for the captured seismic data, vertical component data for the second velocity data in the third data domain;
• transform, from the third data domain to the first data domain, the vertical component data and thereby generate temporal-spatial data associated with the first velocity data;
• generate, based on the temporal-spatial data, the noise component data included in the captured seismic data by subtracting the temporal-spatial data from the pressure data;
• estimate, in a first iteration, the signal component data in the second data domain based on the noise component data, the pressure data, and the vertical component data and thereby generate an estimated signal component data;
• and
• generate a multi-dimensional report indicating the estimated signal component data in the first data domain, the multi-dimensional report indicating the captured seismic data with one of:
• no data values corresponding to the noise component data, or
• substantially attenuated data values of the noise component data.

A computer program for removing noise component data comprised in seismic data, the computer program including a non-transitory computer-readable medium comprising code configured to:

• receive captured seismic data from one or more sensors deployed about a resource site, the captured seismic data including signal component data and the noise component data;
• extract, in a first data domain, velocity data and pressure data included in the captured seismic data and thereby generate first domain data including first pressure data and first velocity data in the first data domain, wherein the first data domain includes a 3-dimensional transform or a 5-dimensional transform, and wherein the first data domain includes a temporal spatial domain;
• transform, from the first data domain to a second data domain, the first velocity data and the first pressure data in the first data domain and thereby generate second domain data including second pressure data and second velocity data in the second data domain, wherein the second data domain includes a Fourier wavenumber domain, wherein transforming, from the first data domain to the second data domain, the first velocity data and the first pressure data comprises performing a Fourier transform along a spatial direction of the first data domain to generate common wavenumber gathers;
• map wavenumber gather data associated with the second pressure data and the second velocity data from the second data domain to a third data domain, wherein the third data domain includes a Fourier-Curvelet domain, and wherein each common wavenumber gather of the Fourier wavenumber domain is mapped to a 2D Curvelet domain;
• apply an enveloping function to the second pressure data and the second velocity data in the third data domain and thereby determine envelope ratio scaling data for the captured seismic data;
• generate, based on noise variance of the second pressure data and the second velocity data, threshold value data for the captured seismic data;
• generate, based on the threshold value data, a thresholding operator for the captured seismic data, wherein the thresholding operator is a stein thresholding operator, a soft thresholding operator, or a combination thereof, and the thresholding operator is applied to the envelope ratio scaling data to generate threshold envelope data;
• apply adaptive localized filtering to the threshold envelope data to generate filter coefficient data from the threshold envelope data for the captured seismic data;
• estimate, based on the coefficient data from the threshold envelope data and a weighted linear combination of the second pressure data and the second velocity data for the captured seismic data, vertical component data for the second velocity data in the third data domain;
• transform, from the third data domain to the first data domain, the vertical component data and thereby generate temporal-spatial data associated with the first velocity data;
• generate, based on the temporal-spatial data, the noise component data included in the captured seismic data by subtracting the temporal-spatial data from the pressure data;
• estimate, in a first iteration, the signal component data in the second data domain based on the noise component data, the pressure data, and the vertical component data and thereby generate an estimated signal component data;
• and
• generate a multi-dimensional report indicating the estimated signal component data in the first data domain, the multi-dimensional report indicating the captured seismic data with one of:
• no data values corresponding to the noise component data, or
• substantially attenuated data values of the noise component data.