While seismic attributes are sensitive to lateral changes in geology, they are also sensitive to lateral changes related to noise. Seismic data are usually contaminated by both random and coherent noise, even when the data have been migrated reasonably well and are multiple free. Certain types of noise can be addressed by the interpreter through careful structure-oriented filtering or post-stack footprint suppression. While seismic attributes are particularly effective at extracting subtle features from relatively noise-free data, if the data are contaminated with multiples or are poorly focused and imaged due to poor statics or inaccurate velocities, the data need to back to the processing team to alleviate those problems.
The preconditioning of seismic data was used to improves the signal-to-noise (S/N) ratio of the seismic data by removing spatial noise or enhancing the coherency and alignment of the reflection events, without unnecessary smoothing or smearing of the discontinuities. Although interpreter usually think of removing unwanted features that effect on seismic attributes results.
The quality of any seismic attributes analysis is dependent on the signal to noise ratio in the data, and the ability to detect events as close to the limit of the seismic resolution as possible. Therefore the first step in the workflow consists of structurally oriented noise cancellation to improve the signal to noise ratio, and spectral enhancement to improve the resolution of thin events.
Seismic data must be edited for seismic attributes analysis, spectral analysis assist in determination amount of noises effect in our data and can give actual information about frequencies and amplitude relations. In this case study, starting by spectral analysis for the all 3D seismic cube, it used to determine random noises frequencies and dominant frequency 20 Hz (Figure 4.1). Then, Band-pass filter designed to enhance signal to noise ratio where it cut the lower and higher frequencies related to noises effect. Band-pass filter applied to enhance data continuity and remove high and low frequencies noise to increase signal to noise ratio (Figure 4.2).
Figure 4.2: Seismic line before bandpass filter (left), seismic line after bandpass filter.
After seismic cube displayed in 3D view it is observed there are channel and it does not appear in amplitude slices but it is very clear in similarity at same time slices, random noise effect on similarity attributes result because similarity attributes algorithm sensitive to any change in seismic wave form, post stack processing enhance signal to noise ratio (Figure 4.3).
Acquisition footprint is an undesirable artifact that masks the geologic features or amplitude variations seen on time slices from the seismic data, especially at shallow times. Mean filter overcome foot print noises and increase reflectors continuity, Figure 4.4 shows different between same inline before and after applying mean filter.
After applying bandpass filter the continuity increase and reduce high frequency noise, Figure 4.5 shows comparison between seismic lines after and before data preconditioning, there are very clear enhancement in continuity of reflector, Applying a mean filter to any line will increase the continuity of reflector and enhance indicating faults and other geological features.
Figure 4.4: Seismic section without data preconditioning (right), seismic section after data preconditioning (left).
Dip variances attributes used to test data preconditioning results before and after noise cancelation, Figure 4.5 shows effect of data preconditioning workflow on dip variance attribute and compare it before seismic noises removing and after canceled noises, where left time slice indicate dip variance attributes before data enhancement in this slice difficult to determine channels where in right slice represent dip variance after data enhancement.
Dip variances sensitive to foot print noises because it depend on difference between the smoothed (DMS) dip of maximum similarity and the local value of the dip of maximum similarity. Random noise effect on continuity of reflectors and scatter DMS results, foot print noises reduce seismic reflectors continuity that effect on dip scan in maximum seismic wave form similarity direction, any dip attributes are sensitive to foot print noises
Figure 4.5: Dip variance time slice without data preconditioning (left), dip variance time slice after preconditioning (right).
Similarity attribute used to detect channels edges and faults identification, similarity attribute applied to test data preconditioning workflow, similarity attributes depend on waveform comparing in vertical time window, random noise effected on seismic wave form, Figure 4.6 shows effect of data preconditioning workflow on similarity attributes result in same time slice where left slice indicate similarity attributes before data preconditioning in this slice difficult to determine channels where in right slice represent dip variance after data enhancement. Seismic attributes helped to determine best seismic data preconditioning parameters suitable for enhance channel interpretations, more smooth seismic data less resolution for thin geological features.
Figure 4.6: Similarity time slice without data preconditioning (right), similarity attributes with data preconditioning (left) showing enhancement in channel edge detection
4.2 generating Seismic Attributes for Identifying Stratigraphic Features
There are too many duplicate attributes, too many attributes with obscure meaning, and too many unstable and unreliable attributes. They are purely mathematical quantities, and they are not really attributes at all. These unnecessary seismic attributes can be identified through inspection aided by crossplots, histograms, and correlation. Discarding redundant and useless attributes leaves a much-reduced set of attributes that is easier to use (Barnes, 2007).
This repetition confusion and makes it hard to apply seismic attributes effectively. You do not need them all. Review your seismic attributes and reduce them to a much smaller subset. Discard duplicate and dubious attributes, prefer attributes with intuitive geologic or geophysical meaning, understand resolution, distinguish processes from attributes, and avoid poorly designed attributes. The subset remaining is both more manageable and more honest (Barnes, 2007).
In this case study, envelope attribute generated to determine channels trend, Figures 4.7 shows envelope attribute extraction in time 0.988 s before smooth mean filter applying so it is difficult to determine channels trend because data is noisy, in (Figure 4.7) left image represents envelope after applying mean filter smoothing that help in determine channel trend and it is observed channel have low envelope amplitude values
Figure 4.7: Envelope attribute time slice before data preconditioning (left), envelope attribute time slice after data preconditioning with enhancement channel image (right).
Average energy is the square of the RMS amplitude, it used to identify channel and study how average energy can help to identify channels after apply data conditions and remove noise effect average energy help to identify channels Report this
- Moses Ilesanmi
- Ioana Silvia MaricaStatus is online
- Syed Daud Shah
- Hossam WalyStatus is reachable
- Maisara YussufStatus is reachable
- Ashraf SamakStatus is reachable
- Omar RashidStatus is reachable
- Mohammed Al-Amrani
- Kamran Laiq, (P.Geo.) (PMP)®Status is reachable
- mohamed ouabelStatus is reachable