Method and apparatus for seismic noise reduction
Methods, apparatuses, and systems are disclosed that reduce seismic noise. In one embodiment, a method of processing seismic data includes accessing seismic data representative of a plurality of seismic input traces acquired by one or more seismic sensors. The method also includes stacking the plurality of seismic input traces into a stacked trace. The method also includes generating, utilizing at least one processor unit, a function of similarity between at least two of the plurality of seismic input traces. The method also includes scaling at least one of the seismic input traces or the stacked trace with the function of similarity.
1. A method for reducing noise in geophysical data processing and imaging of acquired seismic data, the method comprising:
accessing seismic data representative of a plurality of seismic input traces acquired by one or more seismic sensors, wherein each of the plurality of seismic input traces is recorded at a seismic sensor;
stacking the plurality of seismic input traces into a stacked trace;
generating, utilizing at least one processor unit, a function of similarity between at least two of the plurality of seismic input traces, wherein the function of similarity comprises a plurality of respective measurements of similarity generated from a plurality of covariance matrices measuring the similarity of the seismic input traces over a respective plurality of sample windows;
calculating the similarity between the respective plurality of sample windows in the at least two of the plurality of seismic input traces and corresponding sample windows in the stacked trace, wherein the sample windows are selected to differentiate noise from signal in the seismic data;
scaling at least one of the seismic input traces or the stacked trace with the function of similarity, wherein scaling with the function of similarity reduces the noise;
applying an imaging condition to the scaled stacked trace or input traces to generate image data representing a portion of the Earth's subsurface; and
displaying the image data representing a portion of the Earth's subsurface on a display device coupled to the processor unit.
2. The method of claim 1 , further comprising storing data representative of at least one of the scaled at least one seismic input trace, the scaled stacked trace, or a derived property from the scaled at least one seismic input trace or the scaled stacked trace in a computer readable tangible storage medium.
3. The method of claim 1 , further comprising the image data being representative of at least one of the scaled at least one seismic input trace, the scaled stacked trace, or a derived property from the scaled at least one seismic input trace or the scaled stacked trace.
4. The method of claim 1 , wherein the function of similarity is a function of a sample domain of the seismic input traces, and wherein the sample domain is a time domain or a depth domain.
5. The method of claim 1 , wherein the respective measurements of similarity are generated by:
selecting a first sample window of the plurality of sample windows;
calculating a first measurement of the similarity for the first sample window;
sliding the first window by at least one sample to select a second sample window of the plurality of sample windows; and
calculating a second measurement of the similarity for the second sample window.
6. The method of claim 1 , wherein at least one of the respective measurements of similarity is generated responsive to similarities between all of the plurality of input traces.
7. The method of claim 1 , wherein at least one of the respective sample windows is approximately three times a dominant wavelength of an anticipated seismic event.
8. The method of claim 1 , wherein a size of at least one of the plurality of sample windows is designed to isolate an anticipated seismic event.
9. The method of claim 1 , wherein the respective plurality of sample windows in the at least two of the plurality of seismic input traces form a first plurality of sample windows, and the corresponding sample windows in the stacked trace form a second plurality of sample windows.
10. The method of claim 9 , wherein at least one of the plurality of measurements of similarity comprises a mean derived from a covariance matrix for a respective one of the first plurality of sample windows and a corresponding respective one of the second plurality of sample windows.
11. The method of claim 9 , further comprising:
generating the respective covariance matrix for each of the respective first and second pluralities of sample windows using the at least two of the plurality of input traces; and
generating the respective measurements of similarity by determining a mean of each respective covariance matrix.
12. The method of claim 11 , wherein the mean is determined by summing all elements of each respective covariance matrix and dividing a resulting sum by a number of the elements in each respective covariance matrix.
13. The method of claim 11 , wherein elements of the covariance matrix for each respective sample window are generated by the equation
C
ij
=
1
N
-
1
∑
k
=
1
N
(
t
ki
-
t
k
_
)
(
t
kj
-
t
k
_
)
where C ij denotes a covariance in a sample window between a seismic input trace i and a seismic input trace j, t ki is a k th sample in seismic input trace i, N is a number of samples in the sample window, and t k is a k th sample in the stacked trace.
14. The method of claim 1 , wherein the stacked trace is scaled with the function of similarity and further comprising applying the imaging condition to the scaled, stacked trace after it has been scaled with the function of similarity, such that scaling with the function of similarity reduces the noise prior to imaging the data.
15. The method of claim 1 , wherein at least one of the seismic input traces is scaled with the function of similarity and said stacking of the seismic input traces is done after said scaling of the seismic input traces, such that scaling with the function of similarity reduces the noise prior to imaging the data.
16. A method of reducing noise in geophysical data processing and imaging for acquired seismic data, the method comprising:
correcting a plurality of seismic input traces acquired by one or more seismic sensors, wherein each of the plurality of seismic input traces is recorded at a seismic sensor;
stacking the corrected plurality of seismic input traces into a stacked trace;
generating a plurality of covariance matrices measuring the similarity of the corrected seismic input traces over a respective plurality of sample windows, wherein the sample windows are selected to differentiate noise from signal in the seismic data;
generating a plurality of mean values, each mean value representative of a plurality of elements of each respective covariance matrix, each mean value corresponding to one of the plurality of sample windows;
combining the plurality of mean values into a function of similarity;
scaling the stacked trace with the function of similarity, wherein scaling the stacked trace with the function of similarity reduces noise on the scaled, stacked trace as compared with the unscaled, stacked trace; and
storing data representative of at least one of the scaled at least one seismic input trace, the scaled stacked trace, or a derived property from the scaled at least one seismic input trace or the scaled stacked trace in a computer readable tangible storage medium, wherein an imaging condition is applied to the scaled stacked trace or input traces to generate image data representing a portion of the Earth's subsurface; and
displaying the image data on a display device, wherein the noise is reduced prior to imaging the data.
17. A system for reducing noise in geophysical data processing and imaging of acquired seismic data, the system comprising:
a storage device for storing seismic data representing a plurality of seismic input traces acquired by one or more seismic sensors, wherein each of the plurality of seismic input traces is recorded at a seismic sensor;
a processor unit coupled to the storage device and configured to:
stack the plurality of seismic input traces into a stacked trace;
generate a function of similarity between at least two of the plurality of seismic input traces;
scale at least one of the seismic input traces or the stacked trace with the function of similarity, wherein scaling with the function of similarity reduces noise in the seismic data; and
generate image data representative of at least one of the scaled at least one seismic input trace, the scaled stacked trace, or a derived property from the scaled at least one seismic input trace or the scaled stacked trace, wherein the function of similarity comprises a plurality of respective measurements of similarity, and wherein respective measurements in the function of similarity are generated from a plurality of covariance matrices measuring the similarity of the corrected seismic input traces over a respective plurality of sample windows;
calculate the similarity between the respective plurality of sample windows in the at least two of the plurality of seismic input traces and corresponding sample windows in the stacked trace, wherein the sample windows are selected to differentiate the noise; and
a display device coupled to the processor unit and configured to display the generated image data, wherein an imaging condition is applied to the scaled stacked trace or input traces to generate image data representing a portion of the Earth's subsurface, wherein the noise is reduced prior to imaging the data.
18. The system of claim 17 , wherein each respective value in the function of similarity is a mean of each respective covariance matrix.
19. The system of claim 17 , wherein elements of the covariance matrix for each respective sample window are generated by the equation
C
ij
=
1
N
-
1
∑
k
=
1
N
(
t
ki
-
t
k
_
)
(
t
kj
-
t
k
_
)
where C ij denotes a covariance in a sample window between a seismic input trace i and a seismic input trace j, t ki is a k th sample in seismic input trace i, N is a number of samples in the sample window, and t k is a k th sample in the stacked trace.
20. The method of claim 16 , wherein elements of the covariance matrix for each respective sample window are generated by the equation:
C
ij
=
1
N
-
1
∑
k
=
1
N
(
t
ki
-
t
k
_
)
(
t
kj
-
t
k
_
)
,
where C ij denotes a covariance in a sample window between a seismic input trace i and a seismic input trace j, t ki is a k th sample in seismic input trace i, N is a number of samples in the sample window, and t k is a k th sample in the stacked trace.