tSpecial section: Seismic attributes · We have used coherence, curvature, and spectral decomposition seismic attributes to determine the morphol-ogy and gray-level co-occurrence

Seismic attributes for description of reef growth and channel systemevolution — Case study of Intisar E, Libya

Yichuan Wang1, Christoph Georg Eichkitz2, Marcellus Gregor Schreilechner2, Gabor Heinemann3,John C. Davis3, and Mohammed Gharsalla4

Abstract

A 3D seismic survey over the Intisar E field in the Ajdabiya Trough of the Sirte Basin, Libya, revealed achannel-like feature in Eocene carbonates that wraps around the pinnacle reef that contains the reservoir.We have used coherence, curvature, and spectral decomposition seismic attributes to determine the morphol-ogy and gray-level co-occurrence matrix attributes to define seismic facies within the feature. These indicatedthat the channel originated by submarine scouring caused by downslope movement of turbidity currents. Ero-sion was followed by the deposition of successive layers of carbonate debris in the channel. Stratigraphic cor-relations with the adjacent pinnacle reef revealed that the channel was cut during the late stage of reef growth,and a second channel formed after the Intisar E reef ceased to grow. Differences in seafloor elevation over thereef probably diverted turbidity currents so channels were not cut into the reef, breaching the reservoir. Thisinterpreted geologic history may explain why some pinnacle reefs in the Intisar complex contained giant res-ervoirs, whereas others were barren.

IntroductionA 3D seismic survey over the Eocene Intisar E pin-

nacle reef in the Sirt Basin of Libya shows a subtle fea-ture adjacent to the reef that may be a submarinechannel. Seismic attributes were used to define thegeometry of the feature and to interpret its internalcomposition. Numerous investigators such as Chenand Sidney (1997), Marfurt et al. (1998), Chopra andAlexeev (2006), Chopra and Marfurt (2007), and Eich-kitz et al. (2015) have defined different 3D seismicattributes. However, few published reports discussthe use of multiple attributes for detailed interpretationof sedimentary structures in carbonates.

Baaske et al. (2009) describe similar channels in Eo-cene carbonates of Libya and suggest that they could beregional guides in petroleum exploration. We focus onthe specific interpretation of the internal structure ofsediments in the Intisar E channel and the possible in-fluence of the channel on the reservoir in the adjacentIntisar pinnacle reef. Coherence, curvature, and spec-tral decomposition were used for structural interpreta-tion and delineation of channel morphology. Attributesbased on the gray-level co-occurrence matrix (GLCM)helped to distinguish seismic facies.

Geologic background of the case studyThe study area is located in the Sirte Basin of Libya, a

structural depression of approximately 189;000 mi2

(490;000 km2) in extent that was formed as a triple-junction rift along the northern margin of the Africancontinental plate (Ahlbrandt, 2001). Beginning in theLate Cretaceous, the northern edge of the Africancontinental plate was covered by a marginal sea thatpersisted until the Late Eocene (Priabonian). Themarine environment consisted of several broad, shal-low carbonate platforms separated by deeper waterin subsiding troughs.

A tectonic overview of Sirte Basin showing regionalstructural highs and basin centers is given in Figure 1.The green circle indicates the general location of thestudy area, which includes approximately 16 mi2

(43 km2) around the Intisar E oil field. The reservoirconsists of a Paleocene pinnacle reef, and it is partof a series of pinnacle reefs that grew in an embaymenton the eastern side of the Ajdabiya Trough, along themargin of a carbonate platform.

The stratigraphic column in Figure 2 illustrates thecomplexity of the sedimentary succession in the studyarea and the adjacent Cyrenaica Platform. Beginning in

1University of Saskatchewan, Department of Geological Sciences, Saskatoon, Saskatchewan, Canada. E-mail: [email protected] GmbH, Leoben, Austria. E-mail: [email protected]; [email protected] Oil GmbH, Leoben, Austria. E-mail: [email protected]; [email protected] Oil Company, Tripolis, Libya. E-mail: [email protected] received by the Editor 19 January 2015; revised manuscript received 6 August 2015; published online 7 October 2015. This paper

appears in Interpretation, Vol. 4, No. 1 (February 2016); p. SB1–SB11, 9 FIGS.http://dx.doi.org/10.1190/INT-2015-0017.1. © 2015 Society of Exploration Geophysicists and American Association of Petroleum Geologists. All rights reserved.

t

Special section: Seismic attributes

Interpretation / February 2016 SB1Interpretation / February 2016 SB1

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

the Early Cretaceous, movement of the Africancontinental plate created extensive rifts, including theAjdabiya Trough. Syndepositional faulting of rift mar-gins resulted in rapid lateral facial changes. Episodesof marine transgression and erosion are registered asseven regional and local unconformities (Figure 2).

During the Paleocene, the location of active riftingmoved far to the east but regional subsidence contin-ued, providing accommodation on an extensive marineshelf, where a sequence of carbonates more than 500 m(1600 ft) thick was deposited. The lowermost Paleo-cene unit was the Lower Sabil Formation of Danian

age, which unconformably overlies theUpper Cretaceous Kalash Formation.The Lower Sabil Formation is composedof interbedded limestones and shalesthat were deposited in a quiet, shallowwater, low-energy shelf and lagoonalenvironment, possibly behind a barrierreef (Hallett, 2002), whereas to the east,the unit becomes increasingly dolomitic(Bezan, 1996). The overlying SheteratFormation is marine shale, representinga brief marine transgression and drown-ing of the carbonate shelf of the easternSirte Basin in Middle Paleocene (Selan-dian) time. The Sheterat Formation sep-arates the Lower Sabil Formation fromthe basal dolomite-anhydrite section ofthe Upper Sabil Formation. Carbonateshelf conditions returned during theThanetian, and the Upper Sabil succes-sion was deposited, along with its lateralequivalent successions including the in-terfingering of carbonate complexes ofthe Zelten Formation, the Intisar EShale, and the mixed carbonates andshales of the Kheir Formation (Figure 2).These formations are a result of cyclicalvariations in water depth on the shelfduring an overall transgression to deepmarine conditions (Bebout and Pen-dexter, 1975). This interval includesthe carbonate bioherms of the Intisarpinnacle reefs, classified as the UpperSabil Formation, and the enclosingKheir Formation calcareous shales thatact as a seal over the reefs. Upper Sabilshelf-edge deposition was not influ-enced by rifting, and no significant faultsare present in the study area (Rusk,2001). A sequence of shelf carbonatesand evaporates more than 4000 m(13,000 ft) thick was deposited duringthe Eocene, as the Gir, Gialo , and AugilaFormations. The Ypresian-age Gir For-mation consists of basal restricted shelffacies overlain by an evaporite intervalfollowed by the massive carbonates de-posited in an open marine environment.

Seismic attribute methodologyTo illuminate reef bodies, channel

structures, and seismic facies withinthe channel features, we applied several

Figure 1. Tectonic map of Libya showing regional highs and basin centers. Theproject area, located in the Ajdabiya Trough, is indicated by a red square (modi-fied after Rusk, 2001).

SB2 Interpretation / February 2016

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

seismic attributes. For the definition of the channeledges and the reef edges, we applied coherence mea-surements (Blumentritt et al., 2003) and curvatureattributes (Oyedele, 2005). In addition, we applied spec-tral decomposition to describe the channel features andalso to highlight the reef structure. For the descriptionof seismic facies, we used several different seismicattributes, such as spectral decomposition and GLCM-based attributes.

Coherence measures the similarity between the seis-mic waveforms along neighboring traces. Low coher-ence, caused by rapid changes in theseismic waveform between adjacenttraces, can be interpreted as resultingfrom faults, channels, or other abruptdisruptions (Marfurt et al., 1998).

The most common methods for cal-culating coherence include crosscorre-lation (Bahorich and Farmer, 1995),semblance (Marfurt et al., 1998), andeigenstructure analysis (Gersztenkornand Marfurt, 1999). Crosscorrelationand eigenstructure coherence respondto differences in the reflector wave-forms of adjacent traces, whereas sem-blance-based coherence is sensitive toboth lateral changes in reflector ampli-tudes and waveforms. Semblance-basedcoherence was used for all coherencecalculations in this study. The sem-blance σðt; p; qÞ is defined as the ratioof the energy of the average trace tothe average energy of all the tracesalong a specific dip (Marfurt et al.,1998). This provides a measure of thesimilarity between traces within theanalysis window.

Curvature is a measure of the appar-ent bending of seismic reflections andenhances subtle folds, flexures, and col-lapse features that are not easily seen bycoherence (Al-Dossary and Marfurt,2006). In two dimensions, curvature isdefined as the radius of a circle tangentto a curve. In three dimensions, cur-vature is defined by two orthogonalcircles. There are numerous poststackcurvature attributes, including meancurvature, Gaussian curvature, maxi-mum curvature, and others (Roberts,2001). Chopra and Marfurt (2007) pointout that the most-negative kneg andmost-positive curvatures kpos are theeasiest measures to visually correlatewith features of geologic interest.

Spectral decomposition is widelyused for imaging and mapping bed thick-ness and geologic discontinuities from3D seismic data (Partyka et al., 1999).

The amplitude response at different frequencies canbe tuned to a specific bed thickness, which helps tohighlight stratigraphic features such as channels andareas of complex faulting (Torrado et al., 2014). Spec-tral decompositions based on the short-windowdiscrete Fourier transform (SWDFT) and the continu-ous-wavelet transform (CWT) were used in this study.The SWDFT procedure allows an analysis of the fre-quency component in a specific fixed-time window, rep-resenting a subsurface interval at a specific depth(Peyton et al., 1998; Partyka et al., 1999). In CWT analy-

Figure 2. Stratigraphic column for the eastern Sirte Basin and the CyrenaicaPlatform (modified after Burwood et al., 2003). Geologic ages, in millions ofyears, are from the International Chronostratigraphic Chart 2014/02. The greenrectangle encloses the stratigraphic interval examined in this study. Arrows in-dicate the source rock (green), reservoir (yellow), and seal (red).

Interpretation / February 2016 SB3

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

ses, the frequency bandwidth depends on the chosencenter frequency because the bandwidth increases athigher center frequencies (Sinha et al., 2009). TheSWDFT limits the resolution of the time-frequencyspace through a predefined window length, whereasthe CWT does not require the window length in ad-vance. Hence, there is no fixed resolution throughthe time-frequency space for CWT. Wavelet dilationand compression of CWT efficiently offer the optimalwindow length instead of the subjective choice (Sinhaet al., 2005). Thus, the CWT procedure has better res-olution than SWDFT.

The results of spectral decomposition are usually dis-played by RGB (R = red, G = green, and B = blue) blend-ing. In this visualization method, amplitudes at threedifferent frequencies are plotted together using thethree colors. Compared to standard displays, RGBblending creates informative, multiattribute displaysthat show features in finer detail (Henderson et al.2008).

The GLCM is a statistical texture classification meth-odology developed by Haralick et al. (1973) for inter-preting remote sensing images. The GLCM is basedon the frequency of occurrence of different combina-tions of neighboring pixel values in a digital image. Thismethodology has been applied to seismic data for thepast 20 years; a recent summary of the literature isgiven in Eichkitz et al. (2015). The GLCM-based attrib-utes are used primarily to define seismic facies and areapplied here to distinguish channel-filling lithologies.

Seismic attribute interpretation of reef areaThis study is based on a 3D prestack time migration

(PSTM) seismic cube consisting of 231 north–south in-lines and 301 perpendicular crosslines. Geophoneswere spaced at 25-m intervals, and survey lines were25 m apart. The data array covers approximately39 km2. The seismic records were sampled at a 2-msrate.

Based on the first occurrences of the reef structureand the channels, the positions of horizons to be inter-preted were determined. In addition, synthetics andwell top information were used to identify the reflec-tors. After interpretation of all relevant horizons (Fig-ure 3), attribute interpretations were made on ahorizon basis. The objectives of this horizon-basedattribute analysis were to define the reef and channeledges and to describe the channel fillings in more detail.

Seismic attribute interpretation can be based on timeslices or horizon slices. Interpretations of time sliceswere used to locate the beginning of reef growth andthe final formation of the channel system. Then, horizoninterpretations were made of five horizons that corre-spond to key stratigraphic markers. These five horizonswere used for the extraction of seismic attributes alonghorizon slices (Figure 3). The lowest horizon is the topof the Lower Sabil Formation, and it represents the baseof the pinnacle reef. The top of the pinnacle reef corre-sponds to the top of the Upper Sabil Formation. Withinthe pinnacle reef, the seismic response is limited. Inmost parts, chaotic reflections with low amplitudes

can be observed. Above the pinnaclereef are sediments of the Harash andKheir Formations (Figure 3). The inter-preted Kheir horizon includes a subma-rine erosional surface in the form of achannel system that trends approxi-mately north–south within the studyarea. This channel system can also beinterpreted in the Intisar A field. Thechannel is filled with sediments of theGir Formation; these consist of threesuccessive sedimentary intervals infor-mally designated Gir T, Gir R, and Gir S.

We applied semblance-based coher-ence, curvature, spectral decomposition,and textural attributes to describe therelationship between reef growth andchannel evolution. Figure 4 shows aseries of semblance-based coherencetime slices from 1650 to 1870 ms.Growth of the reef and evolution ofthe channel system can be interpretedfrom this series. From the slice at1870–1750 ms, the reef grew and ex-panded. At approximately 1740 ms, thereef body was partially eroded on itseastern side by channel A, whereas onits western side, the reef continued togrow. By Late Thanetian (at approxi-

Figure 3. Crossline showing interpreted horizons Top Kheir (green), Gir T(red), Gir S (light blue), Gir R (dark blue), Gir Q (gray), and Gir P (purple).Top Kheir is interpreted as erosional surface cutting Kheir Formation betweeninlines 2400 and 2500 and partly eroding Harash Formation between inlines 2435and 2500. Gir T, Gir S, and Gir R horizons are interpreted as channel filling.

SB4 Interpretation / February 2016

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

Figure 4. Semblance-based coherence time slices from 1650 to 1870 ms. Blue areas indicate Intisar E reef. Green areas are in-terpreted as channel A and red areas are interpreted as channel B. The reef structure is initially observed at 1870 ms and appears asa circular feature in shallower time slices. Coherence is uniform within the reef body until 1740 ms, when channel A first appearsand bends around the eastern flank of the reef as the western portion continues to grow. At approximately 1710 ms, channel Bappears but does not affect the reef and is absent above 1670 ms.

Interpretation / February 2016 SB5

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

Figure 5. (a) Most-positive curvature attrib-ute displayed on interpreted Top Kheir hori-zon. (b) The area between the right and leftchannel upper margins is color coded to dis-tinguish channels A (green) and B (red).(c) Most-negative curvature attribute. (d)The area between the right and left channellower margins is color coded to distinguishthe two channels. Bar scales indicate curva-ture magnitude.

Figure 6. (a) Seismic coherence attribute of the Top Kheir horizon. (b) The blue color is added to emphasize the extent of channelA. (c) Curvedness combined with the shape index using a 2D colorbar (curvedness versus shape index) shown on the right. Chan-nel margins appear as thin lineaments. (d) Attribute display with colors added to distinguish channel A (blue), channel B (red), andthe reef (green).

SB6 Interpretation / February 2016

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

mately 1730 ms), the reef had stopped growingon the west and sediments of the Harash and KheirFormations were deposited over the reef, where thechannel system was still active during deposition ofthese units. In Late Thanetian to Early Ypresian (at ap-proximately 1710 ms), another channel developed fur-ther east of the reef. This second feature, designated aschannel B, was limited in the vertical and horizontalextents.

Most-positive and most-negative curvature attributesdisplayed on interpreted horizons such as the Top Kheirreveal details of the channel system (Figure 5). Thechannel edges are defined by the most-positive curva-ture attribute and appear as pairs of thin lines ofpositive values (Figure 5a). These represent placeswhere the seismic surface is strongly concave down-ward on the right- and left-hand margins of the channeland indicate the tops of the channelmargins (Figure 5b).The bottom edges of the channel margins are defined bythe most-negative curvature attribute and form a pair oflines of negative values where the surface is stronglyconcave upward (Figure 5c). The areas with low curva-ture values in-between the high curvature channeledges represent the bottom of the channel. The bandformed by the two lines of most-negative curvature

lies within band formed by most-positive curvature(Figure 5).

Other attributes such as coherence express the sur-face form as a statistic that compares the similarity infrequency content of adjacent lines. If a surface con-tains a discontinuity such as a fault or abrupt flexure,the coherence will be low; if the surface is relativelysmooth, the coherence will be high. A display of coher-ence in the Top Kheir horizon is shown in Figure 6a and6b. The margins of the channels are clearly displayed ona coherence gray-scale plot.

Volumetric vector dip estimation can be used to cal-culate additional curvature attributes such as curved-ness, which combines the maximum and minimumcurvatures. The shape index attribute also combinesthe maximum and minimum curvatures, but it doesso in a manner that is scale independent. Curvedness,used in combination with the shape index, can enhancethe edges of discontinuous features. By simultaneouslydisplaying these two attributes using a 2D colorbar, it ispossible to visualize the edges of the channel systems,as seen in Figure 6c and 6d. The distinction betweenchannels A and B is readily apparent.

Spectral decomposition can also be used to distin-guish between the two channel systems because differ-

Figure 7. Spectral decomposition using CWT at (a) 30, (b) 40, and (c) 50 Hz. (d) Frequency cubes of panels (a-c) blendedusing RGB (red: 30 Hz, green: 40 Hz, and blue: 50 Hz). (e) Color coding is added to RGB blended images to distinguish the reefbody from channels.

Interpretation / February 2016 SB7

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

ent frequencies will highlight different features. In Fig-ure 7a–7c, the spectral decomposed cubes for 30 Hz(Figure 7a), 40 Hz (Figure 7b), and 50 Hz (Figure 7c)

are displayed on the Top Kheir horizon. These plotsconfirm that the 30- and 40-Hz components are domi-nant in the seismic data, and that the reef structure con-

Figure 8. (a) The GLCM energy on Top Gir Twith interpretation of Top Kheir shown as abackground. (b) Added red color A indicatesthe counterpoint bar; red color B indicates thepoint bar in channel A; green color C indicatesfill, interpreted as a channel bar, in channel B.(c) The GLCM energy on Top Gir S with inter-pretation of Top Kheir shown as a background.(d) Red color A added to channel A indicatesthe counterpoint bar, whereas the red color Bindicates the point bar in channel A.

Figure 9. Geologic model of the evolution ofpinnacle reef and channel system. (a) Initia-tion of Upper Sabil reef growth, (b) channelA erosion of the eastern part of the reef,(c) continued erosion of channel A and initialerosion of channel B, (d) early Eocenesediments deposited in both channels, and(e) widespread deposition of higher Gir sedi-ments.

SB8 Interpretation / February 2016

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

tains mainly 40-Hz data. The channels and reef structureare highlighted by RGB blending of these three fre-quency cubes (Figure 7d and 7e). The 40-Hz responseof the reef appears in green and contrasts with the chan-nel edges that appear as dark thin lines created by amixture of all three colors. Inside the channels thereare areas where different frequency componentsdominate.

Seismic attributes based on the GLCM were used forinterpretation of the channel filling. With textural attrib-utes such as the GLCM, it is possible to perform a semi-automatic interpretation of seismic facies based on theamplitude values in the seismic data. In this projectarea, three different seismic facies areas have been in-terpreted (Figure 8). Facies B indicates a point bar ofchannel A with a chute channel inside, whereas faciesB is a counterpoint bar. Therefore, facies C is anotherchannel bar of channel B for geologic interpretation.

Based on the combined interpretation of all attrib-utes, a simplified geologic model has been proposedthat describes the evolution of the reef and the associ-ated channel system (Figure 9).

The deposition of the Kheir Formation spanned thetransition from the Paleocene to the Eocene, when arapid, short-lived worldwide transgression occurred,followed by an abrupt regression (Sluijs et al., 2008).This regression resulted in drowning of the carbonateplatform in the Intisar area from the prolific shallowshoal zone into the deeper open marine zone of pelagiccarbonate deposition, inhibiting growth of the Intisarpinnacle reefs and deposition of the Kheir Formationaround and over the reefs. The subsequent rapid regres-sion would have exposed wide swaths of recentlydeposited carbonates, causing their erosion and trans-port toward deeper water in mass movements as sub-marine landslides that became downslope turbiditycurrents (Eberli and Ginsburg, 1988). Such currentsare capable of rapid erosion of the seabed and quicklyscour out channels. The recently deposited fine-grained, argillaceous sediments of the Kheir Formationwould have been especially susceptible to such action.Seismic attribute interpretation shows that multiple ep-isodes of submarine erosion occurred around the Inti-sar E reef. Inside both channels, sediments of the Gir T,Gir S, and Gir R stratigraphic zones were deposited dur-ing the Early Eocene. After the channels were filled,widespread deposition of higher Gir intervals (Gir Q-Gir A) occurred.

ConclusionsThis case study has demonstrated that seismic attrib-

ute analysis can be useful for the interpretation of thegeologic evolution of complex carbonate sequences. Inthis case study, different seismic attributes helped togain an understanding of how and when an Eocene pin-nacle reef started to grow and when submarine chan-nels developed and how they were influenced by thepresence of the reef. Coherence attributes were usedto interpret the channel and reef edges and establish

the temporal relationship between reef growth andchannel evolution. By integrating different seismicattributes, it was possible to describe the evolutionof the channel system. Curvedness and the shape indexwere used in combination to define channel edges andto distinguish between two channels. It could be estab-lished that channel A was formed during the late stagesof reef growth, and the second channel, B, formed dur-ing Late Thanetian to Early Ypresian time after reefgrowth had ceased. During the Early Ypresian, thechannels became filled by sediments of the Gir Forma-tion in at least three episodes of deposition.

Seismic attributes are effective in extracting subtlefeatures, which are sometimes not recognizable onoriginal seismic data. This technology helps acceleratethe seismic interpretation. Besides, we should also rec-ognize the limitation of seismic attributes. Seismicattributes can exacerbate subtle effects such as acquis-ition footprint and velocity pull-up/push-down.

AcknowledgmentsWe wish to thank the Zueitina Oil Company for pro-

viding the data and for permission to publish thisarticle. Critical and constructive reviews by four anony-mous reviewers helped to improve an earlier version ofthis paper.

ReferencesAhlbrandt, T. S., 2001, The Sirte Basin Province of Libya—

Sirte-Zelten total petroleum system: U.S. GeologicalSurvey Bulletin 2202-F, 4–5.

Al-Dossary, S., and K. J. Marfurt, 2006, 3-D volumetricmultispectral estimates of reflector curvature and rota-tion: Geophysics, 71, no. 5, P41–P51, doi: 10.1190/1.2242449.

Baaske, U. P., P. Tricker, J. Van Itterbeeck, H. Griffiths, andJ. Pickens, 2009, Gravity-induced deep-water carbonatedeposits: Potential new plays in the Eocene of the SirteBasin, Libya: Proceedings of the 8th conference on Afri-can E&P, Petroleum Exploration Society of Great Brit-ain/Houston Geological Society, London, 6.

Bahorich, M., and S. Farmer, 1995, 3-D seismic discontinu-ity for faults and stratigraphic features: The coherencecube: The Leading Edge, 14, 1053–1058.

Bebout, D. G., and C. Pendexter, 1975, Secondary carbon-ate porosity as related to early Tertiary depositionalfacies, Zelten field, Libya: AAPG Bulletin, 59, 665–693.

Bezan, A. M., 1996, The Palaeocene sequence in Sirt Basin,in M. J. Salem, A. J. Mouzughi, and O. S. Hammuda,eds., Geology of the Sirt Basin: Elsevier Science B.V.,97–118.

Blumentritt, C. H., E. C. Sullivan, and K. J. Marfurt, 2003,Channel detection using seismic attributes on the Cen-tral Basin Platform, west Texas: 73rd Annual Inter-national Meeting, SEG, Expanded Abstracts, 466–469.

Interpretation / February 2016 SB9

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

Burwood, R., J. Redfern, and M. J. Cope, 2003, Geochemi-cal evaluation of East Sirte Basin (Libya) petroleum sys-tems and oil provenance: Geological Society of London,Special Publications, 207, 203–240, doi: 10.1144/GSL.SP.2003.207.11.

Chen, Q., and S. Sidney, 1997, Seismic attribute technologyfor reservoir forecasting and monitoring: The LeadingEdge, 16, 445–448, doi: 10.1190/1.1437657.

Chopra, S., and V. Alexeev, 2006, Application of textureattribute analysis to 3D seismic data: The Leading Edge,25, 934–940.

Chopra, S., and K. J. Marfurt, 2007, Using curvature to mapfaults, fractures: AAPG Explorer, 28, 42–43.

Eberli, G. P., and R. N. Ginsburg, 1988, Aggrading and pro-grading infill of buried Cenozoic seaways, northwesternGreat Bahama Bank, in A. W. Bally, ed., Atlas of seismicstratigraphy: AAPG, 27, 97–103.

Eichkitz, C. G., M. G. Schreilechner, P. de Groot, and J.Amtmann, 2015, Mapping directional variations in seis-mic character using gray-level co-occurrence matrix-based attributes: Interpretation, 3, no. 1, T113–T23,doi: 10.1190/INT-2014-0099.1.

Gersztenkorn, A., and K. J. Marfurt, 1999, Eigenstructure-based coherence computations as an aid to 3-D struc-tural and stratigraphic mapping: Geophysics, 64,1468–1479, doi: 10.1190/1.1444651.

Hallett, D., 2002, Petroleum geology of Libya: Elsevier Sci-ence B.V.

Haralick, R. M., K. Shanmugam, and I. Dinstein, 1973, Tex-tural features for image classification: IEEE Transac-tions on Systems, Man, and Cybernetics, 3, 610–621,doi: 10.1109/TSMC.1973.4309314.

Henderson, J., S. J. Purves, and G. Fisher, 2008, Delinea-tion of geological elements from RGB color blendingof seismic attribute volumes: The Leading Edge, 27,342–350, doi: 10.1190/1.2896625.

Marfurt, K. J., R. L. Kirlin, S. L. Farmer, and M. S. Bahorich,1998, 3-D seismic attributes using a semblance-based coherency algorithm: Geophysics, 63, 1150–1165,doi: 10.1190/1.1444415.

Oyedele, O., 2005, 3-D high resolution seismic imaging ofdeep water systems, SE Green Canyon, Sigsbee Escarp-ment, Gulf of Mexico: M.S. thesis, University of Houston.

Partyka, G., J. Gridley, and J. Lopez, 1999, Interpretationalapplications of spectral decomposition in reservoircharacterization: The Leading Edge, 18, 353–360, doi:10.1190/1.1438295.

Peyton, L., R. Bottjer, and G. Partyka, 1998, Interpretationof incised valleys using new 3-D seismic techniques: Acase history using spectral decomposition and coher-ency: The Leading Edge, 17, 1294–1298, doi: 10.1190/1.1438127.

Roberts, A., 2001, Curvature attributes and their applica-tion to 3D interpreted horizons: First Break, 19, 85–100, doi: 10.1046/j.0263-5046.2001.00142.x.

Rusk, D. C., 2001, Petroleum potential of the underex-plored basin centers — A twenty-first-century chal-lenge, in M. W. Downey, J. C. Threet, and W. A.Morgan, eds., Petroleum provinces of the twenty-firstcentury: AAPG Memoir 74, 429–452.

Sinha, S., P. Routh, and P. Anno, 2009, Instantaneous spec-tral attributes using scales in continuous-wavelet trans-form: Geophysics, 74, no. 2, WA137–WA142, doi: 10.1190/1.3054145.

Sinha, S., P. S. Routh, P. D. Anno, and J. P. Castagna, 2005,Spectral decomposition of seismic data with continu-ous-wavelet transform: Geophysics, 70, no. 6, P19–P25, doi: 10.1190/1.2127113.

Sluijs, A., H. Brinkhuis, E. M. Crouch, C. M. John, L. Han-dley, D. Munsterman, S. M. Bohaty, J. C. Zachos, G.-J.Reichart, S. Schouten, R. D. Pancost, J. S. S. Damste,N. L. D. Welters, A. F. Lotter, and G. R. Dickens,2008, Eustatic variations during the Paleocene-Eocenegreenhouse world: Paleoceanography, 23, PA4216, doi:10.1029/2008PA001615.

Torrado, L., P. Mann, and J. Bhattacharya, 2014, Applica-tion of seismic attributes and spectral decompositionfor reservoir characterization of a complex fluvial sys-tem: Case study of the Carbonera Formation, Llanosforeland basin, Colombia: Geophysics, 79, no. 5,B221–B230, doi: 10.1190/geo2013-0429.1.

Yichuan Wang received a B.E.(2013) in geological engineering fromthe China University of Mining andTechnology, Beijing. He studied forapproximately one year (2014) atMontanuniversitaet Leoben, Austria.During this period, he worked as a re-searcher at Joanneum Research, Insti-tute of Geophysics and Geothermics

and finished his master’s thesis related to seismic attrib-utes. He is now pursuing a Ph.D. in geophysics from theUniversity of Saskatchewan, Canada. His current researchinterests include elastic impedance, acoustic-impedanceinversion, and amplitude variation with offset.

Christoph Georg Eichkitz receiveda master’s degree (2005) in appliedgeophysics from MontanuniversitaetLeoben, Austria, where his researchwas focused on modeling of micro-gravity data. He cofounded Geo5GmbH in July 2015, where he isresponsible for seismic interpretation,attribute calculation, and structural

modeling. Prior to founding Geo5 GmbH, he worked forJoanneum Research, Institute of Geophysics and Geother-mics for eight years. His main interests include seismicattribute calculation, pattern recognition, and interpreta-tion workflows.

SB10 Interpretation / February 2016

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

Marcellus Gregor Schreilechner re-ceived M.Sc. (1996) and Ph.D. (2007)degrees in applied geosciences fromthe Montanuniversitaet Leoben, Aus-tria. He is the CEO of Geo5, whichis specialized for geophysical ser-vices, research, and development.He lectures at the University of Tech-nology Graz in applied geophysics.

His main interests include the sequence stratigraphic inter-pretation of seismic reflection data, application of seismicattributes for reservoir characterization, and environmen-tal geophysics.

Gabor Ferenc Heinemann receivedan M.S. (1993) in physics from theGraz University of Technology, Aus-tria and a Ph.D. (2005) in reservoir en-gineering from MontanuniversitätLeoben, Austria. Since 2003, he hasbeen the head of Heinemann OilGmbH, an oil and gas consulting Com-pany in Leoben, Austria. He has exten-

sive experience in oil and gas management, reservoirconsulting and engineering, and software engineering.He has participated as an engineer and a project managerin more than 25 major studies worldwide. He is the authorof many research and technical papers on reservoirsimulation.

John C. Davis received an M.S.(1963) and a Ph.D. (1967) in geologyfrom the University of Wyoming,USA. After a brief stint on the facultyat Idaho State University, he joinedthe University of Kansas where heserved as chief of the mathematicalgeology section of the Kansas Geo-logical Survey and as a professor in

the Department of Chemical and Petroleum Engineering.His textbook, Statistics and Data Analysis in Geology(3rd ed., 2002), leads in its field. After retiring in 2003,he joined the Faculty of Montanuniversitaet Leoben, Aus-tria, where he was a professor of geostatistics and reser-voir characterization until 2006. He is now a chief geologistfor Heinemann Oil GmbH in Leoben.

Mohamed Marbruk Gharsalla re-ceived a B.S. from Elfathe Universityin Tripoli, Libya, and M.S. and Ph.D.from Montanuniversität Leoben, Aus-tria, all in petroleum engineering. Heis a reservoir engineering superin-tendent in the Engineering PlanningDepartment at Zueitina Oil Companyin Tripoli, Libya. Since 2015, he has

also been working with Heinemann Oil Company as aproject manager in oil and gas development. His researchinterests include production optimization, especially fornaturally fractured reservoirs.

Interpretation / February 2016 SB11

Dow

nloa

ded

11/0

8/15

to 1

78.1

89.5

9.15

8. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/