AccScience Publishing / JSE / Article
Submit to JSE
Cite this article
39
Views
Journal Browser
Volume | Year
Issue
Search
Forthcoming Issue
Current Issue
View All
News and Announcements
View All
ARTICLE

Diffraction imaging using specularity gathers

I. STURZU1 ,  A.M. POPOVICI1 ,  T.J. MOSER2
Show Less
1 Z-Terra Inc., 17171 Park Row, Houston, TX 77084, U.S.A. isturzu@z-terra.com,
2 Moser Geophysical Services, van Ikenadelaan 550A, 2597 AV The Hague, The Netherlands.,
JSE 2014, 23(1), 1–18;
Submitted: 30 May 2013 | Accepted: 2 November 2013 | Published: 1 February 2014
© 2014 by the Authors. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution -Noncommercial 4.0 International License (CC-by the license) ( https://creativecommons.org/licenses/by-nc/4.0/ )
Download PDF
Cite
XML
Abstract

Sturzu, I., Popovici, A.M. and Moser, T.J., 2014. Diffraction imaging using specularity gathers. Journal of Seismic Exploration, 23: 1-18. The separate imaging of subsurface diffractors is a key ingredient in the development of high-resolution imaging technologies. We here produce images of diffractors using depth migration algorithms modified to attenuate the energy from specular reflectors. The seismic events from a pre-stack seismic dataset are migrated to proper depth and location using the final velocity model obtained by the velocity model building process, but the output is assigned to separate bins according to the value of a specific parameter called specularity. The specularity gathers are post-processed using a plane wave destructor filter to attenuate the contribution coming from specular reflectors. The method is demonstrated on two synthetic models and on a field data target in the Teapot Dome reservoir.

Keywords
seismic migration
diffraction imaging
high-resolution imaging
plane wave destructor
common image gather
specularity gather
References
  1. Furthermore, the high frequencies that are present in the data are often lostduring standard processing. High-resolution imaging is of value, for instance toenable identification of small scale faults and to locate formation pinch-outs.
  2. Standard approaches to obtain high-resolution information, such as coherencyanalysis and structure-oriented filters, derive attributes from stacked, migratedimages. In comparison, diffraction imaging can act directly on the pre-stackdata, and has the potential to focus and image super-resolution structuralinformation as a consequence of the redundancy present in the pre-stack data.THEORY AND METHOD
  3. Diffractions are the seismic response of small elements (or diffractors) inthe subsurface of the earth, such as small scale faults, near surface scatteringobjects, and, in general, all objects which are smaller than the seismicwavelength. Diffraction imaging uses diffractions to focus and image thestructural elements that produced those diffraction events. Since diffractors are,by definition, smaller than the wavelength of seismic waves, diffraction imaginghas the potential of providing super-resolution information, to image details thatare beyond the classical Rayleigh limit of half a seismic wavelength. Theimportance of diffractions in high-resolution structural imaging has beenemphasized in many recent publications (Shtivelman and Keydar, 2004;
  4. Khaidukov et al., 2004; Taner et al., 2006; Fomel et al., 2006; Moser and
  5. Howard, 2008; Moser, 2009; Klokov et al., 2010; Klokov et al., 2011; Delland Gajewski, 2011; Koren and Ravve, 2011; Klokov and Fomel, 2012). Still,diffraction imaging is not widely used tool. In fact, most algorithms that areused to process seismic data explicitly enhance reflections and implicitlysuppress diffracted energy. The goal of diffraction imaging is not to replacethese traditional algorithms, but rather to provide an additional 3-D or 4-Dvolume containing information about diffractors that would be able to fill in thesmall, but potentially crucial, structural details.DIFFRACTION IMAGING 3
  6. A true diffraction image is not optimally obtained by post-processing atraditional seismic image even if the seismic image is obtained by an algorithmthat does not suppress diffractions. While diffractors will appear in the image,usually in the form of discontinuities, they have much lower amplitudes thanreflecting structures. On the other hand, by imaging diffractors using thepre-stack data, the diffractor amplitude can be enhanced while the specularreflections can be attenuated. Furthermore, apparent discontinuities in theseismic image can have a variety of causes other than diffractions, includingsmall errors in the velocity model of the earth that was used to obtain the image,so the post-stack/post-processing approach would not be able to discriminatebetween discontinuities coming from diffractions and those coming fromprocessing errors.
  7. Techniques for diffraction imaging fall into two categories. In the firstcategory are methods that separate the seismic data into two parts, one thatcontains the wave energy from reflections (specular energy) and the other thatcontains the wave energy from diffractions. Each component is used to providean image through traditional seismic imaging methods. In the second categoryare methods that do not separate the input seismic data, but rather performsfiltering during migration. Moser and Howard (2008) and Moser (2009)extracted the local direction of specularity from a previously obtained migrationstack, and used this information during a subsequent migration step in order tofilter the events that satisfied (to a given degree) Snell’s law. Koren and Ravve(2011) pre-computed a directivity-dependent specularity attribute usinginformation from the velocity model and the acquisition geometry and usedangle domain gathers in order to suppress the specularity energy associated withhorizontal events in the angle domain gathers. In this paper we follow theapproach from Moser and Howard (2008). The parameters governing thespecularity filtering are rather arbitrary, if no further investigation is implied.
  8. Sturzu et al. (2013) introduced a new concept - specularity gathers - that provedto be very useful in the design of proper parameters for the specularity filter.
  9. Below we show how we can selectively filter the specular energy withinspecularity gathers to obtain the diffraction images after stacking along thespecularity dimension.
  10. In Kirchhoff migration, energy is propagated to all possible reflectionpoints in the model space. After all events on all traces are propagated, animage is generated by stacking (summing) all individual contributions. Thepropagation of the events usually uses Green’s functions computed in the formof travel-time tables (the time of propagation from the source defined by thetrace to the image point and further to the receiver defined by the trace.)
  11. Stacking reinforces in-phase energy corresponding to true reflectors and cancelsout-of-phase energy that does not correspond to a true reflector. A conventionalKirchhoff migration forms a seismic image as:4 STURZU, POPOVICI & MOSERV(x) = ) dtdsdrU'(t,s,r)é6[t 一 T(s,x,r)] , (1)where 6 is the Dirac delta function, U'(t,s,r) the (second time derivative)pre-stack data, depending on time t and shot/receiver positions s/r, T(s,x,r) isthe travel time from s to r via the subsurface image point x, computed by raytracing in a given reference velocity model, and V(x) the resulting migratedimage. The sum is carried out over the time samples and all source and receiverpairs (s.r), in the seismic data.
  12. In the earth’s subsurface the local discontinuities (reflectors) can bemodeled either as surfaces, edges, or isolated tips or points. For a smoothsurface reflector, the corresponding part of the image is a locally continuous,planar surface that generates specular events, meaning that they strictly obey
  13. Snell’s law. Events backscattered from all other types of discontinuities arediffractive and do not obey Snell’s law. Diffraction imaging attenuates thecontribution of specular events in a migrated image.
  14. In the final image stack, a specular element can be approximated locallyas a planar surface, while the isochrone surfaces (computed using the travel-timetables) should be tangent to the planar surfaces corresponding to the specularelements, as a direct consequence of Snell’s law. For diffractive events,however, there is no such constraint. For a given trace in the data and a pointin the image (1), we define the specularity in terms of the specularity angle,defined by the normal to a locally planar structure (dominant at that imagepoint) and the direction of the gradient of the total travel-time from the sourceto the receiver via the image point. We choose the absolute value of the cosineof the specularity angle as the actual value of specularity.. This can be expressedmathematically as:S(.xr) = [wT |/|T I, Q)where T, denotes the gradient of T(s,x,r)with respect to x, and n is the unitvector normal to the reflector surface, also depending on x, and the dot denotesthe scalar product. If x is located on a strong reflector, the value of thespecularity should be equal to unity, S = 1, because the two rays (coming fromthe source and that going toward the receiver) then obey Snell’s Law withrespect to the normal to the reflector n, so the angle bisector of the two rays isaligned with the normal to the surface. In diffraction imaging framework, thisis a pure specular reflection that has to be attenuated. If the angle bisector of therays and the normal to the surface are not perfectly aligned, S < 1 and theenergy is non-specularly scattered.
  15. The concept of pure specularity as defined above has to be amended usingthe concept of the Fresnel zone, which is a frequency-dependent volume aroundDIFFRACTION IMAGING 5the ray within which most of the wave energy is interfering constructively andcan be treated as a single arrival wave. All the points from a Fresnel zone haveto be considered together, even if they have slightly sub-unitary values forspecularity. That is why pure specularity has to be defined by a frequency-dependent interval close to unity. In Fig. 1, we illustrate this situation bycomparing a pure specular reflection with a pure diffraction. The Fresnel zonesare depicted by hatched areas.
  16. Fig. 1. Comparison between a pure specular reflection and a pure diffraction case. The Fresnel zonefor the reflection are depicted by the hatched areas.
  17. A straightforward procedure for obtaining a diffraction image is outlinedin Moser and Howard (2008) and Moser (2009). First, using pre-stack Kirchhoffmigration, we obtain the seismic image; this image will include both reflectionsand diffractions, but the reflections dominate the image. The second step is toanalyze the structures in the Kirchhoff image and determine the normal vectorto these structures at each image point. Using a migration stack to obtain thisinformation, rather than to extract it from the final velocity model, is critical incases when the stratigraphic non-conformity is important (Moser, 2009). Thesecond step has to be performed on an optimally focused image, obtained usingthe best velocity model, so that the information extracted is related to thegeological geometry of the undersurface. In a subsequent migration run, themigrated seismic events are stacked using a weight designed in order toattenuate the contribution of the specular events (specularity value close to 1).6 STURZU, POPOVICI & MOSER
  18. An important challenge comes from the fact that there is no a prioriprocedure to define the limits of the pure specularity region as a function ofspecularity itself, so one cannot design a proper weighting function before thelast migration run. A simple trial-and-error method can be too computationallydemanding. Specularity gathers, a novel technique introduced in Sturzu et al.(2013) can be used to increase the efficiency and accuracy of the diffractionimaging technique. A specularity gather is similar to an offset or angle commonimage gather, in which the migrated seismic events are separated according tothe value of specularity rather than of the offset or, respectively, reflectionangle. The events are migrated to the proper depth and are partially stackedaccording to the specularity values in pre-defined specularity bins. Thespecularity gather can be formally written as:
  19. V,.(x,8) = J dtdsdrU'(t,s,r)6[t — T(s,x,n)]6(S — |n-T,|/T,|]) . @)
  20. In a post-processing technique similar to the muting of the offset gathers,the diffraction image can be obtained after a weighted stack over all thespecularity values:Vux) = J dSwix,S)V, (8) . 4)
  21. The use of specularity gathers has the advantage that the weightingfunction is designed after migration and therefore is constructed, and updated,very efficiently. In particular, the weighting function can be spatially variable[w = w(x,S)] and adapted to the local Fresnel zone width, which is difficult toestimate a priori, but becomes feasible using specularity gathers. Also, feedbackfrom interpretation can be easily included in the weighting function, and hencein the final diffraction image.
  22. As shown in Sturzu et al. (2013), for a correct velocity model and in theideal infinite-frequency limit, a specular reflection event appears in thespecularity gathers as a focused spot on the S=1-axis. Point diffractions appearas flat events extending over 0 < S < 1. Edge diffractions in three dimensionsappear as dipping events, as they obey Snell’s law only along the edge, but nottransversely to it (Moser, 2011). For finite bandwidth seismic responses, thesituation is slightly different. Specular reflections also appear as dipping events,as the non-specular part of reflected energy outside the Fresnel zone is notrelated to the shortest reflection path following Fermat’s principle.
  23. In a Common Image Point section of the specularity gather, the image isobtained from different portions of the isochrones of different traces. Forexample, in the case of a horizontal flat reflector with a constant velocityDIFFRACTION IMAGING 7overburden, the isochrones are ellipses. The specularity angle in each point ofthe ellipse is monotonically changing from zero below the Common Image Pointto 90 degrees at zero depth. In a Common Image Point section of the specularitygather, the bin for the maximum specularity (specularity angle close to zero) isformed by pieces of the ellipses corresponding to maximum depth. Here, traceswith any offset should contribute. The very next specularity bin (in the same
  24. Common Image Point section) is formed by the contributions of the traces withneighboring mid-points, but having the right offset to yield the designed valueof the specularity. The location of the events should be shallower than thelocation of the event in the previous bin. This interpretation pattern can beapplied to subsequent specularity bins, ending with the bin for zero specularity,which has to have contributions only at zero depth, but from all traces.
  25. Fig. 2 shows migration results for the case of a horizontal flat reflectorwith a constant velocity overburden. The sub-figures depict: (a) a Common
  26. Image Point (or vertical) section from the specularity gather (a partial imageobtained for a given vertical in the image space and all values of thespecularity), (b) the final stack over all values of specularity, equivalent to thestandard migrated image, and, in each of panels (c), (d) and (e) a specularitysection (a partial image obtained for a given value of the specularity for allpoints in the image space) for S = 1.0, S = 0.9 and S = 0.8, respectively.
  27. Sub-Figs. 2(d) and 2(e) contain out of Fresnel zone ghosts for the main specularevent depicted in sub-Figs. 2(b) and 2(c). Remarkably, stacking over all valuesof specularity is able to fully cancel the contribution of all the ghosts in the finalHorizontal rane an
  28. Fig. 2. Migration results for a horizontal flat reflector with a constant velocity overburden: (a) thespecularity gather for the horizontal position at x = 1250 m; (b) The stack over all values ofspecularity, equivalent to a standard migrated image; (c) specularity section for S = 1.0; (d)specularity section for S = 0.9; (e) specularity section for S = 0.8.Depth [m]8 STURZU, POPOVICI & MOSERstack (b). Fig. 3 show similar results for a point diffractor in a constant velocitymedium. In the left panels is displayed the migration image, obtained bystacking along all specularity values in the specularity gather. The specularitygather in the exact location of the point diffractor is a flat horizontal event(sub-figure 3(a) central panel); when moving slightly away from this point, thehorizontal event splits into two (sub-Fig. 3(b) central panel). The right panelsshow two specularity sections, for S = 1.0 in sub-Fig. 3(a) and for S = 0.8 insub-Fig. 3(b).
  29. Horizontal position [m] Specularity Horizontal position [m]500 100015002000 0s 1.0 500 100015002000500 os 10 500
  30. Fig. 3. Migration results for a point diffractor in a constant velocity medium. Each panel displaysfrom left to right: the final stack, the specularity gather for a given horizontal position, x, and aspecularity section for given value of specularity, S. The horizontal location of the diffractor is at1250 m. (a) x= 1250 m, S = 1.0 (b) x = 1150 m, S = 0.8.a)Depth fm1b)Depth [m]DIFFRACTION IMAGING 9
  31. Displaying common-image specularity gathers (vertical sections) maybecome cumbersome when dealing with more complicated data. Fortunately,using a common-depth display, i.e., showing sections along one of thehorizontal lines (corresponding to depth in the common-image gather) versusspecularity on the vertical axis is able to give a clearer image, especially forcases with small lateral variations. For these cases, in the Common-Depth
  32. Specularity Gathers (horizontal sections), the specular reflections are almosthorizontal events. An important issue is that in this display one can identify (outof Fresnel zone-) ghosts of the specular events coming from deeper locations,which are also almost horizontal. In this way, we can filter these ghosts togetherwith their primaries.
  33. A workflow for diffraction imaging using common-depth specularitygathers consists of:
  34. I Standard pre-stack depth migration using formula (1) and associatedmigration velocity analysis to obtain an optimally focused full-wave imageV(x);
  35. II. Extraction of the unit vector normal to the reflector surface using V(x) ineach point;
  36. Ill. Migrating using eq. (3) to obtain a specularity gather;
  37. IV. Filtering the specular energy from the specularity gather;
  38. V. Stacking over specularity dimension to obtain a diffraction image [eq.(4)].NUMERICAL RESULTS
  39. The numerical results are obtained using the procedure outlined above[eqs. (1)-(4)]. The filtering step IV can be done using any procedure able todetect and attenuate laterally continuous seismic events. Here we used one ofthem, the Plane Wave Destruction Filter (PWD) (Fomel, 2002). Before applyingthe filter, we compute the dips in each section of the Common-Depth
  40. Specularity Gather. Then, in each point from the gather, the filter is performinga weighted stack along the dip in order to attenuate the seismic event from agiven vertical window, if a similar event is found along the dip. The firstnumerical example has been designed for a proof of concept (Fig. 4) andillustrates the functionality of specularity gathers on a simple diffraction rampmodel: a horizontal reflector at 900 m of depth, and a double ramp with thebase at 1400 m, as depicted in the perfect migration stack shown in the panel(a) from Fig. 4. The synthetic data are generated using ray-Born approximation,10 STURZU, POPOVICI & MOSERa method proved to be very useful in forward modeling diffracted waves(Moser, 2012). After pre-stack migration, the unit vector normal to thereflectors in the final stack (a) was computed and used to generate a specularitygather [eq. (3)] in a subsequent migration run. The Common-Depth SpecularityHorizontal position [m] Horizontal position [m]上和上上(b)一o—Depth{m]SSpecularity0 0.0 Os 10=一=区SpecularityOs0 0.0(h)&Specularity0s0
  41. Fig. 4. Diffraction ramp model: (a) Pre-stack migration image obtained by stacking over the valuesof specularity in the specularity gather. (b) Diffraction image obtained by stacking over specularityof the Plane Wave Destructor (PWD) filtered specularity gather. (c) Section of the specularity gatherin common-depth display for 1400 m of depth. (d) Section of the specularity gather in common-depthdisplay filtered with PWD for 1400 m of depth. (e) Section of the specularity gather incommon-depth display for 1140 m of depth. (f) Section of the specularity gather in common-depthdisplay filtered with PWD for 1140 m of depth. (g) Section of the specularity gather incommon-depth display for 900 m of depth. (h) Section of the specularity gather in common-depthdisplay filtered with PWD for 900 m of depth.DIFFRACTION IMAGING 1
  42. Gather of Fig. 4c shows two horizontal events close to S = 1 coming from thespecular reflections shown in the stack from Fig. 4a at 1400 m depth, while forthe diffractive events from the same depth at 750 m, 1500 m, and 2250 m alongthe line, there are clearly defined peaks. Close to the central peak, we noticealso two dipping events from the specular reflections close to the edge of thedouble-ramp. After applying the Plane Wave Destruction Filter the specularenergy is attenuated, and the result showing three diffraction peaks is displayedin Fig. 4d. The specularity gather of Fig. 4e does not have horizontal events,but displays, close to S = 1, dipping events corresponding to the specularreflections at the top end of the ramps, visible in the stack from Fig. 4a at 1140m in depth. For the diffractive events from the same depth at 750 m and 2250m along the line, there are clearly defined peaks. In Fig. 4f is displayed theresult of applying the Plane Wave Destruction Filter on the section from Fig.4e: the specular energy is almost completely attenuated and two diffractionpeaks are visible. The specularity gather of Fig. 4g shows a horizontal eventclose to S = 1 coming from the specular reflections shown in the stack from
  43. Fig. 4a at 900 m in depth, and two ghost dipping events coming from thespecular reflections on the ramps. After filtering with the Plane Wave
  44. Destruction Filter, we obtain almost no energy - except for two very weak peaksat the survey’s edges - as displayed in Fig. 4(h). Stacking the filtered specularitygathers over the values of specularity gives the diffraction image shown in Fig.4(b). Almost all of the specular energy was attenuated in the final image leavingjust the five points of discontinuity in the model.Horizontal position [m] Specularity4000 8000 12 000 0.0 0.5 1.0(a) 8(b)SpecularityDepth [m]
  45. Fig. 5. Mare di Cassis model: (a) Specularity gather in common-depth display (horizontal section)for 2100 m. (b) Specularity gather in common-image display (vertical section) for the horizontalposition x = 4880 m.12 STURZU, POPOVICI & MOSER
  46. The second example is the Mare di Cassis data set, which is described in
  47. Moser and Howard (2008). After regular pre-stack migration, the unit vectornormal to the reflectors in the final stack was computed and used to generate aspecularity gather in a subsequent migration run. Fig. 5 displays a comparisonbetween a Common Depth Specularity Gather (horizontal section) on the left,and a vertical section from the specularity gather on the right. Visually it isclear that the first one displays more information: the specular reflections areidentified as laterally continuous events, the out-of-Fresnel zone ghosts areidentified as similar events at smaller values of specularity, while diffractionsare identified by the numerous peaks. In the vertical display the diffractions areidentified by the horizontal peaks, while specular reflections by dipping eventsthat tend to align toward the zero-specularity point from the surface. Thespecularity gathers were tapered along the specularity axis, and the planewave-destructor filter was applied in the common depth sections of the gather.Horizontal position [m]6 000 8 000 10 000 12 000(a) 8Specularity(b)Specularity
  48. Fig. 6. Mare di Cassis model: (a) Common-Depth Specularity Gather for 1310 m. (b) PWD filteredCommon-Depth Specularity Gather for 1310 m.DIFFRACTION IMAGING 13
  49. In Figs. 6, 7, and 8, the result of applying the plane wave-destructor filteron a common depth specularity gather is shown respectively for three values ofdepth. Almost all the specular energy is attenuated, except for regions close to
  50. S=1 where the dip calculation is affected by either multiple dips concurrentlyin the same image point or by vertical dips; consequently, the planewave-destructor filter is not able to clean all the specular energy. Fig. 9 displaysin panel (a) the standard migrated image, in panel (b) the diffraction imageobtained by stacking over specularities smaller than 0.97 of the planewave-destructor filtered specularity gather, and - for reference - in panel (c), thediffraction image obtained using a cubic taper to filter out all the events fromthe specularity gather corresponding to values of specularity larger than 0.92(Sturzu et al., 2013). A visual comparison shows that the current proceduregives better results than that obtained using the uniform taper.Horizontal position [m]0 2000 4000 6000 8000 10 000 12 000
  51. Fig. 7. Mare di Cassis model: (a) Common-Depth Specularity Gather for 1960 m. (b) PWD filteredCommon-Depth Specularity Gather for 1960 m.14 STURZU, POPOVICI & MOSERHorizontal position [m]6 000 8000 10 000Specularity(b)Speculerity
  52. Fig. 8. Mare di Cassis model: (a) Common-Depth Specularity Gather for 2080 m. (b) PWD filteredCommon-Depth Specularity Gather for 2080 m.
  53. The third example contains a field dataset from Teapot Dome (Powder
  54. River Basin, Wyoming). Here, diffraction imaging has been carried out usingthe same steps as above, but in the framework of a full 3D depth imagingprocess. Figs. 10 and 11 show the results for a region that contains the targetknown as the Tensleep formation. In Fig. 10 we focus on the crossline 118 ofthe survey: panel (a) is the standard migration result (the top of the Tensleepformation is depicted in the figure), while panel (c) displays correspondingsection from the diffraction image. The vertical section of the specularity gatherat the location given by the vertical thin white line on the stack is shown in Fig.10b, while the corresponding section from the plane wave-destructor filteredspecularity gather is displayed in Fig. 10d. The filtering procedure was appliedin each horizontal section (common depth specularity gather) separately, so Fig.10d was not obtained by applying directly the plane wave-destructor in thevertical section.DIFFRACTION IMAGING 15Horizontal position [m]a 4000 6 000 8 000 12 00010 000Depth [m](b)Depth [m](c)Depth [m]
  55. Fig. 9. Mare di Cassis model: (a) Standard migrated image; (b) Diffraction image obtained bystacking over specularity of the PWD filtered specularity gather; (c) Diffraction image obtained usinga uniform taper filter above S = 0.92.16 STURZU, POPOVICI & MOSER
  56. In Fig. 11, we focus on a depth section at 2020 m (depicted in Fig. 10a with athin white horizontal line). Fig. 11a displays the depth section through thestandard migration result, while in Fig. 11b is shown the corresponding depthsection in the diffraction image. The common depth specularity gather for thecrossline 118 at the same depth is shown in Fig. 11c, while the correspondingresult filtered with plane wave-destructor method is depicted in Fig. 11d. In thiscase, due to the 3D geometry, the diffraction peaks from the specularity gatherare not as clear as in the synthetic examples. However, applying the proceduredescribed above is able to delineate in the migrated image (Fig. 11b)high-resolution diffractive elements related to the transition between differentstratigraphic formations.(a) (b)0 2000001600 2 400人全已N(©) (d)
  57. Fig. 10. Teapot Dome dataset, crossline 118: (a) Standard migrated image; (b) Vertical section ofthe specularity gather at inline 138; (c) Diffraction image; (d) Vertical section of the specularitygather at inline 138 filtered with PWD.CONCLUSIONS
  58. Specularity gather analysis proves to be a very useful instrument inobtaining and/or optimizing diffraction images. The energy corresponding tohigher values of specularity can be attenuated by using tapers (uniform or basedon interpretation input). An automatic algorithm can be alternatively constructedby using a filter, such as plane-wave-destructor, to attenuate the specular energyat any location in the specularity gather. Further development of this methodwill help in advancing diffraction imaging technology.DIFFRACTION IMAGING 17(b)Crosslines0 150 100 50Specularity5加 — 全
  59. Fig. 11. Teapot Dome dataset, depth section at 2020 m: (a) Standard migrated image; (b) Diffractionimage; (c) Common-Depth Specularity Gather, display at crossline 118; (d) Common-Depth
  60. Specularity Gather filtered with PWD, displayed at crossline 118.ACKNOWLEDGEMENTS
  61. We thank Rocky Mountain Oilfield Testing Center and the U.S.
  62. Department of Energy for providing the Teapot Dome dataset, and Opera (Pau)for providing the Mare di Cassis dataset. Special thanks are due to ouranonymous reviewer, for comments that greatly improved the manuscript.REFERENCES
  63. Dell, S. and Gajewski, D., 2011. Common-reflection-surface-based workflow for diffractionimaging. Geophysics, 76: 187-195.
  64. Fomel, S., Landa, E. and Taner, M.T., 2006. Post-stack velocity analysis by separation and imagingof seismic diffractions. Expanded Abstr., 76th Ann. Internat. SEG Mtg., New Orleans:2559-2563.
  65. Fomel, S., 2002. Applications of plane-wave destruction filters. Geophysics, 67: 1946-1960.
  66. Khaidukov, V., Landa, E. and Moser, T.J., 2004. Diffraction imaging byfocusing-defocusing: An outlook on seismic superresolution. Geophysics, 69: 1478-1490.
  67. Klokov, A., Baina, R., Landa, E., Thore, P. and Tarrass, I., 2010. Diffraction imaging for fracturedetection: synthetic case study. Expanded Abstr., 80th Ann. Internat. SEG Mtg., Denver:3354-3358.
  68. Klokov, A., Baina, R. and Landa, E., 2011. Point and edge diffractions in three dimensions.Extended Abstr., 73rd EAGE Conf., Vienna: B023.
  69. Klokov, A. and Fomel, S., 2012. Separation and imaging seismic diffractions using migrateddip-angle gathers. Geophysics, 77: 1-S143.
  70. Koren, Z. and Ravve, I., 2011. Full-azimuth subsurface angle domain wavefield decomposition andimaging, Part I: Directional and reflection image gathers. Geophysics, 76: S1-S13.18 STURZU, POPOVICI & MOSER
  71. Landa, E., Fomel, S. and Reshef, M., 2008. Separation, imaging, and velocity analysis of seismicdiffractions using migrated dip-angle gathers. Expanded Abstr., 78th Ann. Internat. SEGMtg., Las Vegas: 2176-2180.
  72. Moser, T.J. and Howard, C.B., 2008. Diffraction imaging in depth. Geophys. Prosp., 56: 627-641.
  73. Moser, T.J., 2009. Diffraction imaging in subsalt geometries and a new look at the scope ofreflectivity. Expanded Abstr., EAGE Subsalt Imaging Workshop: Focus on Azimuth, Cairo.
  74. Moser, T.J., 2011. Edge and tip diffraction imaging in three dimensions. Extended Abstr., 73rdEAGE Conf., Vienna.
  75. Moser, T.J., 2012. Review of ray-Born modeling for migration and diffraction modeling. StudiaGeophys. Geodaet., 56: 411-432.
  76. Shtivelman, V. and Keydar, S., 2005. Imaging shallow subsurface inhomogeneities by 3D multipathdiffraction summation. First Break, 23: 39-42.
  77. Sturzu, I., Popovici, A.M., Tanushev, N., Musat, I., Pelissier, M.A. and Moser, T.J., 2013.
  78. Specularity gathers for diffraction imaging. Extended Abstr., 75th EAGE Conf., London:We-01-03.
  79. Taner, M.T., Fomel, S. and Landa, E., 2006. Separation and imaging of seismic diffractions usingplane-wave decomposition. Expanded Abstr., 76th Ann Internat. SEG Mtg., New Orleans:2401-2404.
  80. Zeng, H., 2009. How thin is a thin bed? An alternative perspective. The Leading Edge, 28:1192-1197.
Share
Back to top
Journal of Seismic Exploration, Electronic ISSN: 0963-0651 Print ISSN: 0963-0651, Published by AccScience Publishing
We use cookies on our website to ensure you get the best experience.
Read more about our cookies here
Accept