Separation of seismic sources by joint interpolation and deblending
Generally, seismic data may provide valuable information with regard to the description such as the location and/or change of hydrocarbon deposits within a subsurface region of the Earth. The present disclosure generally discusses techniques that may be used by a computing system to analyze a data set including weak-coherence signals (e.g., non-coherent blending noise). In particular, a computing system may detect portion of the weak-coherence signals of a gather due to the overlap of selected seismic source excitations and use a mask to isolate coherent signals and the other weak-coherence signals from the masked portion of weak-coherence signals. The coherent signals and other weak-coherence signals may be iteratively processed and used to predict values of the masked weak-coherence signals.
1 . A method of seismic exploration above a region of subsurface containing structural or stratigraphic features conducive to determining a presence, migration, or accumulation of hydrocarbons, the method comprising:
accessing at least a portion of a blended seismic source survey containing blended seismic data, wherein the blended seismic source survey corresponds to at least two interfering seismic source excitation groups produced by at least two seismic source groups having different signatures or frequency characteristics;
producing unblended seismic data by iteratively performing an interpolation and deblending operation on the blended seismic data, comprising:
estimating both first unblended signals and second unblended signals different from the first unblended signals, from the blended seismic data by applying a mask to the blended seismic data to filter from the first unblended signals cross-talk from any seismic source group of the at least two seismic source groups unrelated to the first unblended signals and to filter from the second unblended signals cross-talk from any seismic source group of the at least two seismic source groups unrelated to the second unblended signals; and
predicting at least a portion of the weak-coherence signals caused by the at least two interfering seismic source excitation groups based on a portion of the blended seismic source survey in which the least two interfering seismic source excitation groups overlap above a threshold amount; and
producing at least two source gather groups based on the unblended seismic data to be used to identify hydrocarbons within the region of the subsurface and determine a path along which to drill into Earth to explore for the hydrocarbons.
2 . The method of claim 1 , wherein the iterative interpolation and deblending operation comprises:
selecting a first coherency constraint for a first seismic source group of the at least two seismic source groups;
selecting a second coherency constraint for a second seismic source group of the at least two seismic sources, wherein the second coherency constraint is different from the first coherency constraint; and
applying the first coherency constraint and the second coherency constraint during the deblending operation to the portion of the blended seismic source survey, wherein producing the at least two source gather groups is based at least in part on applying the first coherency constraint and the second coherency constraint.
3 . The method of claim 2 , wherein the threshold amount corresponds to an amount of overlap which permits at least the first coherency constraint and the iterative interpolation and deblending operation to be used to separate at least a portion of data for the first seismic source group from at least an additional portion of data for the second seismic source group.
4 . The method of claim 1 , wherein the iterative interpolation and deblending operation comprises:
indicating at least one weak-coherence portion of a trace associated with the blended seismic source survey;
removing from the trace the at least one weak-coherence portion of the trace; and
predicting a value for the at least one weak-coherence portion of the trace based at least in part on at least one coherent portion of the trace.
5 . The method of claim 4 , wherein the at least one weak-coherence portion of the trace is indicated via null-defining symbol, character, or value stored within the mask.
6 . The method of claim 4 , wherein predicting the value for the at least one weak-coherence portion of the trace comprises referencing at least two data set portions logically disposed adjacent to the at least one weak-coherence portion.
7 . The method of claim 1 , wherein the iterative interpolation and deblending operation comprises applying interpolation operations to:
predict signals in a uniform grid; and
reconstruct signals back to an original non-uniform grid based on the predicted signals in the uniform grid, when the signals are acquired in the non-uniform grid.
8 . The method of claim 1 , wherein the at least two source gathers respectively comprise seismic data representative of a different frequency range and at least some predicted data corresponding to the weak-coherence signals generated based at least in part on the iterative interpolation and deblending operation.
9 . The method of claim 1 , wherein the iterative interpolation and deblending operation comprises:
determining at least one weak-coherence portion of a trace associated with the blended seismic source survey;
generating a mask to logically indicate a location of the at least one weak-coherence portion of the trace within a data set corresponding to the trace;
removing the at least one weak-coherence portion of the trace from the data set corresponding to the trace; and
combining at least one predicted value for the at least one weak-coherence portion of the trace based at least in part on at least one coherent portion of the trace.
10 . The method of claim 9 , wherein removing from the at least one weak-coherence portion of the trace from the data set comprises applying the mask to the data set to zero-out the at least one weak-coherence portion of the trace from the data set.
11 . The method of claim 1 , wherein the at least two seismic source groups comprise a combination of: vibrational sources, airgun sources, sparker sources, and low-frequency sources.
12 . The method of claim 1 , wherein producing the unblended seismic data by iteratively performing the interpolation and deblending operation on the blended seismic data comprises applying a combing operator to masked seismic data obtained by applying the mask to the blended seismic data.
13 . The method of claim 12 , wherein the combing operator comprises a transpose of a blending operator of the interpolation and deblending operation.
14 . The method of claim 1 , wherein the mask corresponds to overlapping seismic signals from interfering seismic sources of the at least two seismic source groups.
15 . A tangible, non-transitory, machine-readable media, comprising instructions configured to cause a processor to:
access at least a portion of a blended seismic source survey, wherein the blended source seismic survey containing blended seismic data and that corresponds to at least two interfering seismic source excitation groups, wherein the seismic source excitations are produced by at least two seismic source groups comprising a first source group and a second source group;
produce unblended seismic data by iteratively performing an interpolation and deblending operation on the blended seismic data in which the processor is caused to:
estimate both first unblended signals and second unblended signals different from the first unblended signals, from the blended seismic data by applying a mask to the blended seismic data to filter from the first unblended signals cross-talk from any seismic source group of the at least two seismic source groups unrelated to the first unblended signals and to filter from the second unblended signals cross-talk from any seismic source group of the at least two seismic source groups unrelated to the second unblended signals; and
predict at least a portion of weak-coherence signals caused by the at least two interfering seismic source excitation groups based on a portion of the blended seismic source survey in which the least two interfering seismic source excitation groups overlap above a threshold amount;
produce at least two separate source gather groups based at least in part on the unblended seismic data, wherein the at least two separate source gather groups are respectively associated with the data sets that are non-overlapping or overlapping within the same source excitation group; and
transmit the at least two separate source gathers for use in generating a seismic image, wherein the seismic image represents hydrocarbons in a subsurface region of Earth or subsurface drilling hazards and is used to identify the hydrocarbons within the subsurface region of Earth and determine a path along which to drill into the Earth to explore for the identified hydrocarbons.
16 . The tangible, non-transitory, machine-readable media of claim 15 , wherein the at least two seismic sources each respectively transmit seismic source excitations comprising different signatures or frequency characteristics, and wherein the at least two seismic sources comprise a combination of one or more of: vibrational sources, airgun sources, sparker sources, or low-frequency sources, or any combination thereof.
17 . The tangible, non-transitory, machine-readable media of claim 15 , wherein the instructions for producing the unblended seismic data by iteratively performing the interpolation and deblending operation on the blended seismic data comprise instructions configured to cause the processor to:
determine at least one weak-coherence portion of a trace associated with the blended seismic source survey;
generate the mask to logically indicate a location of the at least one weak-coherence portion of the trace within a data set corresponding to the trace;
remove the at least one weak-coherence portion of the trace from the data set corresponding to the trace; and
combine at least one predicted value for the at least one weak-coherence portion of the trace based at least in part on at least one coherent portion of the trace.
18 . The tangible, non-transitory, machine-readable media of claim 17 , wherein the instructions for removing the at least one weak-coherence portion of the trace from the data set comprise instructions configured to cause the processor to:
zero-out the at least one weak-coherence portion of the trace from the data set at least in part by applying the mask to the data set.
19 . The tangible, non-transitory, machine-readable media of claim 15 , wherein the instructions for producing the unblended seismic data by iteratively performing the interpolation and deblending operation on the blended seismic data comprise instructions configured to cause the processor to:
indicate at least one weak-coherence portion of a trace in a data set associated with the blended seismic source survey;
remove from the data set the at least one weak-coherence portion of the trace;
predicting a value for the at least one weak-coherence portion of the trace based at least in part on the iterative interpolation and deblending operation being performed on the remaining portions of the trace in the data set; and
combining the remaining portions of the trace in the data set with the predicted value of the at least one weak-coherence portion in a separate source gather corresponding to the first source group.
20 . The tangible, non-transitory, machine-readable media of claim 19 , wherein the at least one weak-coherence portion of the trace corresponds to a time at which seismic source excitations of the first source group interfered with seismic source excitations of the second source group.
21 . The tangible, non-transitory, machine-readable media of claim 19 , wherein transmitting the at least two source gather groups comprising transmitting the separate source gather despite an initial presence of the at least one weak-coherence portion of the trace.
22 . A device, comprising:
a memory storing instructions; and
a processor coupled to the memory and configured to execute the instructions, which cause the processor to:
access at least a portion of a blended seismic source survey containing blended seismic data, wherein the blended source seismic survey corresponds to at least two interfering seismic source excitation groups, wherein the seismic source excitations are produced by at least two seismic source groups comprising a first source group and a second source group;
produce unblended seismic data by iteratively performing an interpolation and deblending operation on the blended seismic data in which the processor is caused to:
estimates both first unblended signals and second unblended signals, different from the first unblended signals, from the blended seismic data by applying a mask to the blended seismic data to filter from the first unblended signals cross-talk from any seismic source group of the at least two seismic source groups unrelated to the first unblended signals and to filter from the second unblended signals cross-talk from any seismic source group of the at least two seismic source groups unrelated to the second unblended signals; and
predicts at least a portion of weak-coherence signals caused by the at least two interfering seismic source excitation groups based on a portion of the blended seismic source survey in which the least two interfering seismic source excitation groups overlap above a threshold amount;
produce at least two source gather groups based on the unblended seismic data, wherein the at least two source gather groups are respectively associated with the data sets that are non-overlapping or overlapping within the same source excitation group; and
transmit the at least two source gathers for use in generating a seismic image, wherein the seismic image represents hydrocarbons in a subsurface region of Earth or subsurface drilling hazards and is used to identify the hydrocarbons within the subsurface region of Earth and determine a path along which to drill into the Earth to explore for the identified hydrocarbons.
23 . The device of claim 22 , wherein the processor is configured to produce the unblended seismic data based at least in part on iterative processes that configured to be repeated until an end condition is met.
24 . The device of claim 22 , wherein the processor is configured to produce the unblended seismic data at least in part by:
determining at least one weak-coherence portion of a trace associated with the blended seismic source survey;
generating a mask to logically indicate a location of the at least one weak-coherence portion of the trace within a data set corresponding to the trace;
removing the at least one weak-coherence portion of the trace from the data set corresponding to the trace by applying the mask to the data set; and
combining at least one predicted value for the at least one weak-coherence portion of the trace based at least in part on at least one coherent portion of the trace.