SeismicStatGraph.pdf

t

Special section: Interpreting stratigraphy from geophysical data

Whither seismic stratigraphy? Bruce S. Hart

/ rg o . g es . y r ar b l/i :/ p tt h esat U f o s rm e T ee s ;t h g ir y p o c r o es en ilc G E S o t ct je

1

Abstract Here, I provide an historical summary of seismic stratigraphy and suggest some potential avenues for future collaborative work between sedimentary geologists and geophysicists. Stratigraphic interpretations based on reflection geometry- or shape-based approaches have been used to reconstruct depositional histories and to make qualitative and (sometimes) quantitative predictions of rock physical properties since at least the mid1970s. This data, is theinterest seismic in stratigraphy that is usually by geology-focused interpreters. First applied to 2D seismic seismic stratigraphy was practiced reinvigorated by the development of seismic geomorphology on 3D volumes. This type of reflection geometry/shape-based interpretation strategy is a fairly mature science that includes seismic sequence analysis, seismic facies analysis, reflection character analysis, and seismic geomorphology. Rock property predictions based on seismic stratigraphic interpretations usually are qualitative, and reflection geometries commonly may permit more than one interpretation. Two geophysics-based approaches, practiced for nearly the same length of time as seismic stratigraphy, have yet to gain widespread adoption by geologic interpreters even though they have much potential application. The first is the use of seismic attributes for “feature detection,” i.e., helping interpre ters to identify stratigraphic bodies that are not readily detected in conventional amplitude displays. The second involves rock property (lithology, porosity, etc.) predictions from various inversion methods or seismic attribute analyses. Stratigraphers can help quality check the results and learn about relationships between depositional features and lithologic properties of interest. Stratigraphers also can contribute to a better seismic analysis by helping to define the effects of “stratigraphy” (e.g., laminations, porosity, bedding) on rock properties and seismic responses. These and other seismic-related pursuits would benefit from enhanced collaboration between sedimentary geologists and geophysicists.

u sb n o ti u ib rt si d e R . 1 6 . 8 9 . 4 6 .1 8 9 o t 4 1 / 8 1 / 7 0 d e ad lo n

Introduction Seismic stratigra phy is an approach to seismic interpretation that is based on principles of stratigraphy. The science of seismic stratig raphy was first formalized in a series of papers published in AAPG Memoir 26 in 1977, although it has older roots (Cross and Lessenger, 1988). Papers (e.g., Mitchum et al., 1977 ) published in the AAPG volume showed how reflection terminations, continuity, and other qualitative “geometry-based” (mostly) analyses of reflections could be used to help to infer depositional histories and predict lithology in undrilled areas. This seismic-based approach became an invaluable tool in the petroleum industry ’s exploration and development efforts and subsequently spread to other applied and fundamental geosciences pursuits. Seismic stratigraphic analyses based on 2D seismic data were later expanded to application on 3D data sets.

routinely applied to stratigraphic analyses even if seismic data are not available. Catuneanu (2006) and Miall (2010) summarize these techniques and discuss the historical development of sequence stratigraphy. Nevertheless, seismic stratigraphy and sequence stratigraphy are so genetically linked that many geoscientists consider the two to be essentially synonymous. In this paper and elsewhere (Hart, 2011), I argue for a somewhat distinct definition of seismic stratigrap hy. Textbooks in sedimentary geology define lithostratigraphy as the study of stratigraphic units that are defined on the basis of their lithology, biostratigraphy as the characterization and correlation of sedimentary deposits based on their fossil content, chemostratigraphy as the use of inorganic chemistry as a correlation tool, and so on. In that spirit, seismic stratigraphy should be defined as the study of stratigraphic units that are defined on the basis of their seismic characteristics. This broad

w o D

Seismic stratigraphy helped to spawn the development of sequence stratigraphy, a science that is now

definition is consistent with the definition of Cross and Lessenger (1988) who defined seismic stratigraphy as

1

Statoil, Houston, Texas, USA. E-mail: [email protected]. Manuscript received by the Editor 23 April 2013; published online 8 August 2013. This paper appea rs in I NTERPRETATION, Vol. 1, No. 1 (August 2013); p. SA3 –SA20, 21 FIGS. http://dx.doi.org/10.1190/INT-2013-0049.1. © 2013 Society of Exploration Geophysicists and American Association of Petroleum Geologists. All rights reserved.

Interpretation / August 2013 SA3

“

/ rg o . g es . ry ar b l/i :/ p tt h esat U f o s rm e T ee s ;t h g ir y p o c r o es en ilc G E S o t ct je u sb n o ti u ib rt si d e R . 1 6 . 8 9 . 4 6 .1 8 9 o t 4 1 / 8 1 / 7 0 d e ad lo n w o D

the science of interpreting or modeling stratigraphy, sedimentary facies, and geologic history from seismic reflection data.” As defined here, seismic stratigraphy can be an end in its own right (e.g., the study of how different stratigraphic features manifest themselves seismically) or it can be one part of a multidisciplinary toolkit that is integrated for other applied or fundamental purposes (e.g., basin analysis, sequence stratigraphy, reservoir modeling, geotechnical studies). Seismic stratigraphy can be used to study depositional basins in their entirety (perhaps from 10 4 to >105 km2 in extent), individual hydrocarbon reservoirs (from < 10 to 102 km2 in extent), for geotechnica l site investigatio ns ( ≪1 km2 in extent) and for other purposes. Seismic stratigraphic principles and techniques have even been adapted for stratigraphic analyses of ground-penetrating radar data sets that commonly cover areas much smaller than 1 km2 (e.g., Jol and Bristow, 2003 ). AAPG Memoir 26 is best known in geologic circles for papers that showed how stratigraphic principles could be applied to seismic interpretation, and that approach is summarized in the first part of this paper. However, it is noteworthy that other papers in the collection focused on rock properties, seismic modeling, and complex trace attributes ( Gregory, 1977; Meckel and Nath, 1977 ; Taner and Sheriff, 1977 ). These and other geophysics- and rock physics-based approaches subsequently failed to gain traction with geologyfocused interpreters, although they are areas where stratigraphers and geophysicists share much common interest. Subequent sections of this paper suggest ways that geophysics-based methods can be applied to help with stratigraphic interpretations and how stratigraphers can help geophysicists to better understand relationships between geology and seismic responses 2.

Snedden and Sarg, 2008). Seismic stratigraphy was srcinally developed for the analysis of 2D seismic data because 3D seismic volumes (and associated visualization technologies) did not become widely available until the 1980s and 1990s. As srcinally defined, there were three distinct but related aspects of seismic stratigraphy: 1) Seismic sequence analysis uses reflection termin ations (e.g., downlap, onlap, erosional truncation) on 2D data, or vertical transects throug h 3D volumes, to define key stratigraphic surfaces (e.g., unconformities, maximum flooding surfaces). These surfaces are then used to assign relative stratigraphic ages to portions of the seismic data (e.g.,

2 For stratigraphers and geophysicists to collaborate more effectively, they will need to become more familiar with each other’s lan-

Unit A is older than Unit B) and to help define other aspects of the depositional history (e.g., history of relative sea-level change). Integration of this “uncalibrated” stratigraphic framework with biostratigraphic data from wells gives the seismically defined packages absolute chronostratigraphic significance (e.g., Vail et al., 1984 ). Seismic sequence analysis is the branch of seismic stratigraphy that eventually gave birth to sequence stratigraphy. Seismic sequence analysis may not lead to inferences about changes global (or even relative) sea level or other driving mechanisms if the data sets cover too small an area (e.g., Hart et al., 2007 ), if the data sets are from basinal settings where factors other than sea level control sediment accumulation, or if other data types are not available to calibrate the seismic-based stratigraphy. It should also be noted that, although seismic stratigraphic analyses were srcinally purported to be useful for identifying and correlating global (eustatic) changes in sea level from basin to basin, criticisms of the methods used (e.g., seismic and biostratigraphic resolution problems) have led many, if not most, stratigraphers to shy away from using seismic surfaces as interregional correlation tools (e.g., Miall, 2010 ). 2) Seismic facies analysis. This appro ach uses reflection geometries and amplitudes as seen on vertical transects (e.g., 2D seismic data) to define seismic facies that are linked by the interpreter to specific stratigraphic bodies (e.g., submarine fan lobes, mass transport complexes, leveed channel complexes) that can be used to make qualitative lithology predictions away from existing well control (e.g., Prather et al., 1998; Colombera et al., 2012). The predictions can be made quantitative , at least in a probabilistic way, if analog databases of properties are available (e.g., net:gross ratios of frontal splays in

guage and paradigms. To facilitate this task, Hart (2011) describes geophysical terms and workflows in ways that should be understandable by geologists, and geologic concepts in ways that should be understandable by geophysicists. SEG’s Wiki sit e ( http://wiki.seg .org) provides additional information about geophysical terms, whereas the Society for Sedimentary Geology’s “Strata” site (http:// www.sepmstrata.org) provides descriptions and definitions of stratigraphy terms and concepts.

deepwater fan systems). Calibration of seismic facies with well control enhances confidence in the interpretation because seismic facies are nonunique. For exampl e, a “chaotic” seismic facies can be associated with reef cores in carbonate settings, mass-transport deposits in deepwater clastic settings, and other lithofacies types.

Seismic stratigraphy: Analyses based on reflection geometries As described in the Introduction, papers published in AAPG Memoir 26 presented techniques for making stratigraphy-based interpretations of seismic data. The initia l emphasis in seismic stratig raphy was on the use of reflection geometries (amplitudes were treated descri ptively) as seen in 2D seismic profiles for exploration purposes. The approach became widely accepted, and subsequent work helped refine and promote the science (e.g., Shell Oil Company, 1987 ; Cross and Lessenger, 1988 ; Bertram and Milton, 1996 ;

SA4 Interpretation / August 2013


Figure 1.Seismic reflections approximate timelines and cross lithologic boundaries within the limits of seismic resolution. This

example shows a progradational Cretaceous succession in a 2D seismic profile collected by the United States Geological Survey from the North Slope of Alaska. (a) Uninterpreted seismic profile. Note the vertical scale, in meters, has been approximated using local well control. (b) Profile showing seismic surfaces, continuous seismic reflections that can be traced over most of the length of the profile. Note the well-developed clinoform geometries with topset, foreset, and bottomset portions to the reflections. The seismic surfaces are dashed in places in the lower foreset portions where strata have been disrupted by approximately syn-depositional deformation. The blue dots represent the location of the paleo shelf break, the break in slope between the relatively gently dipping topset reflections and the more steeply dipping foreset reflections. Changes in trajectory of the shelf break (upward, downward, outward) are used to distinguish whether the system is aggrading, prograding, retrograding, or degrading. (c) Profile showing depositional settings represented by the different portions of the clinoforms. Note that the seismic surfaces cut across several different types of depositional setting and so must cross lithological boundaries. (d) Lithostratigraphic interpretation of the profile, wherein the stratigraphic names are assigned on the basis of lithology. The Torok Formation corresponds to marine shales and deep water clastics, whereas the Nanushuk Group consists of shallow-marine and nonmarine clastics (e.g., Houseknecht and Schenk, 2004). The seismic surfaces cut across lithostratigraphic formation boundaries and so are better for defining depositional histories than the diachronous lithostratigraphic units. From Hart (2011). Reprinted by permission of the AAPG, whose permission is required for further use.

3) Reflection character analysis focuses on latera l var-

represent depositional timelines. Although there are ex-

iations in the character of reflections, or groups of reflections, to predict lateral variations in the stratigraphy (lithology, porosity, thickness, etc.). Seismic modeling can play an important role in this pursuit. Modeling of acoustic responses (i.e., reflections associated with compressional waves and density) of stratigraphic features has become less commonplace, with increased emphasis on modeling of elastic respons es (reflections associated with compressional waves, shear waves, and density) because the latter provide additiona l constraints on lithology and physical property predictions.

ceptions (e.g., Tipper, 1993 ; Zeng and Kerans, 2003 ), this has proven to be a useful starting point for most seismic stratigraphic interpretations. The seismic surfaces (seismic sequence boundaries) shown in Figure 1b can therefore be used to divide the seismic transect into units of relative geologic age using the Law of Superposition; the clinoforms at left represent sedimentary deposits that are younger than those at right. Biostratigraphic data from wells, if any, along the seismic transect could be used to assign more rigorous ages. Although not illustrated here, reflection terminations (onlap, downlap, etc.; Figure 2) could be identified on the image in Figure 1 and used to designate some of the seismic surfaces as sequence boundaries or maximum flooding surfaces3. Instead, I have chosen to illustrate how seismic facies analysis of reflection geometries, amplitudes, and other characteristics of the seismic profiles can be integrated with other data sets (wireline

Seismic stratigraph y offers significant advantages over the more traditional lithostratigraphic interpretation of basins, as illustrated in Figure 1. Part (a) shows an uninterpreted 2D seismic line from the North Slope of Alaska. The image shows a series of clinoforms that represent the infill of a Cretaceous foreland basin. In part (b), some of the more continuous reflections have been identified ( “picked”). These reflections represent Bertram and Milton “seismic surfaces ” in the sense of (1996), i.e., they are laterally continuous reflections against which other reflections terminate. One of the fundamental tenets of seismic stratigraphy is that, within the resolution of the seismic data, reflections

logs, outcrop analogs,environments etc., not shown) probable depositional withintotheinterpret seismic sequences (Figure 1c). Note that even this simple 3 The SEPM ’s stratigraphy Web site (http://www.sepmstrata.org) has mapping and sequence stratigraphy exercises that are based on this same public-domain Alaska seismic data set.


/ rg o . g es . ry ar b l/i :/ p tt h esat U f o s rm e T ee s ;t h g ir y p o c r o es en ilc G E S o t ct je

seismic stratigraphic interpretation has more predictive power than the lithostratigraphic interpretation shown in Figure 1d. Integrating the seismic surfaces with the seismic facies analysis is important because it would allow the interpreter to evaluate changes in depositional processes from one seismic sequence to the next. From a reservoir perspective, it is also important because the rocks associated with these seismic “surfaces” can act as barriers or baffles to lateral and vertical fluid flow; the lateral continuity of the toe-ofslope facies (potentially sandy reservoirs ) suggested by Figure 1c is likely to be more apparent than real. The advent of 3D seismic technology led to a fourth branch of seismic stratigraphy called seismic geomorphology (e.g., Posamentier, 2004 ). Once 3D data sets became commercially available, it quickly became apparent that plan-view images derived from 3D seismic volumes could be used to detect and map stratigraphic features (e.g., Brown et al., 1981 ). Time slices, horizon slices, proport ional slices, and other types of plan-view images now are routinely used to generate paleogeographic images from 3D cubes (e.g., Brown, 2004 ; Weimer and Slatt, 2004 ). These images can be very useful for fundamental studies of depositional systems (e.g., Figure 3) and for qualitative rock property predictions based on relationships between depositional elements and properties of interest (see seismic facies discussion above). From a seismic sequence perspective, the ability to view 3D seismic data from many angles is particularly significant because reflection terminations and configurations are most easily observed from specific angles with respect to the stratigraphic bodies being imaged.

u sb n o ti u ib rt si d e R . 1 6 . 8 9 . 4 6 .1 8 9 o t 4 1 / 8 1 / 7 0 d e ad lo n w o D

For example, Figure 4 shows mutually orthogonal slices through a 3D seismic cube that are viewed from different angles. The data set images a Tertiary shelf-phase

Figure 3. Seismic amplitude map of the Cadotte Member, a

Cretaceous-age clastic strandplain depositfrom Alberta, Canada. (a) Uninterpreted map showing prominent amplitude trends that strike approximately east–west, with other trends more north–south in orientation. Note the acquisition geometry is oblique to these trends, suggesting that the amplitudes are not acquisition artifacts. (b) Interpreted map. The east–west-striking amplitude lineations are interpreted to indicate strandplain orientation (i.e., they represent paleoshorelines), whereas the approximately north–south trends are related to: (1) Cretaceous channels that cut through the Cadotte shoreline, (2) Cretaceous Figure 2. Two different representations of stratal termina-

tions that might be visible in seismic data, well-log cross sections, or sometimes exceptional outcrops. (a) Reflection terminations as defined by Exxon workers in AAPG Memoir 26 (Mitchum et al., 1977). Reprinted by permission of the AAPG, whose permission is required for further use. (b) Reflection terminations as defined by BP in 1996 (Redrawn and used with permission from Bertram and Milton, 1996).


channels in the stratigraphic unit immediately above the Cadotte (i.e., the Paddy Member) that interfere with the Cadotte amplitudes, or (3) poor-data areas beneath a modern river floodplain. Seismic modeling, log-based stratigraphic interpretations , and comparison of the amplitude map to modern surficial features were used to constrain the amplitude interpretation. SeeMcCullagh and Hart (2010)for further details.Reprinted by permission of the AAPG, whose permission is required for further use.

/ rg o . g es . ry ar b l/i :/ p tt h esat U f o s rm e T ee s ;t h g ir y p o c r o es en ilc G E S o t ct je u sb n o ti u ib rt si d e R . 1 6 . 8 9 . 4 6 .1 8 9 o t 4 1 / 8 1 / 7 0 d e ad lo n o w D

delta. Reflection termination s showing toplap, downlap, and erosional truncation are more clearly seen in some orientations than in others. Horizontal slices through clinoforms allow true progradation directions to be determined. The 3D seismic visualization software allows the viewer to pick these or other orientations, whereas an interpreter working with 2D seismic data would be constrained to view the data in the orientation in which they were collected. The 3D seismic data sets should also be used by interpreters to help make the conceptual link between the expression of seismic facies seen on vertical transects and in plan (map) views. This exercise can help interpreters to more confidently interpret 2D data. The 3D

several tens of km 2 ) can be much smal ler than the sequences or systems tracts being imaged. In these small data sets, it is commonly impossible to see the reflection termination patterns that are necessary to identify seismic-scale systems tracts. As such, they need to be integrated with regional data sets consisting of longer 2D seismic lines and/or log-derived correlations ( Hart et al., 2007 ).

seismic interpretation packagesstratigraphic also permitfeatures interpreters to detect and visualize in various ways, such as volume (opacity) rendering and geobody detection (Figure 5). These types of extractions and visualizations need to be undertaken in the context of a fit-for-purpose seismic stratigraphic analysis. Some of the large 2D and 3D (covering several 2 1000 km ) seismic data sets collected offshore provide unparalled opportunities for visualizing entire depositional systems (e.g., Saller et al., 2004 ). Conversely, small land surveys (covering areas of several km 2 to

Figure 4. The seismic expression of stratigraphic features in

a 3D seismic cube depends on which way the data volume is sliced. (a) Vertical transect and (b) time-slice images through a prograding deltaic system. Compare the seismic expression

Figure 5.A simple example of geobody detection. (a) A dis-

of the clinoforms and the incised valley from one image to the other. The clinoforms show toplap and downlap in the vertical transect, but the time slice shows them as lineations that can be used to determine the shoreline orientation. The incision in the vertical transect is seen to have a meandering geometry in the time slice. Area of timeslice covers 96 km 2 (∼38 square miles). From Hart (2011) . Reprinted by permission of the AAPG, whose permission is required for further use.

continuous trough (red) is observed in a vertical transect and used as a seed point by the software to track the feature through the 3D data set, only a portion of which is shown here. The software traces the body as a series of connected voxels through the data set (using user-defined thresholds) and, in this case, identifies a channel feature (b and c). From Hart (2011). Reprinted by permission of the AAPG, whose permission is required for further use. Interpretation / August 2013 SA7


Figure 6. In some cases, stratigraphic features that are invisible to P-waves can be detected using S-waves, as shown by these

images from the Cretaceous section of Alberta. (a) Time slice through P-P data ( “conventional” seismic data — P-wave down and P-wave up). The black dots running from top to bottom toward the left side of the image indicate the location of oil wells that produce from channel sands. There is only a faint indication from this P-wave image that a channel might be present. (b) A corresponding time slice through a P-S data volume (i.e., mode-converted S-waves recorded). This image much more clearly indicates the presence of a channel connecting the productive oil wells. From Margrave et al. (1998).


Seismic stratigraphy was srcinally conceived using, and applied to, compressional wave seismic data. However, the methods of seismic stratigraphy also can be applied to multicomponent data, a subdiscipline referred to as “elastic wavefield seismic stratigraphy ” by Hardage and Aluka (2006) . Shear-wave seismic data can sometimes detect stratigraphic features that are invisible to compressional waves. For example, Figure 6 shows two different time slices through some Cretaceo us clastics of the Western Canada Sedimentary Basin in Alberta. The image on the left shows a conventional seismic image obtained by recording (and then processing) P-wave reflections. A channel corresponding to the producing wells (black dots) is, at best, possibly visible in the data. The converted-wave data4 (right side of Figure 6) more clearly delineate the extent of the channel because the channel-filling deposits respond to shear waves more clearly than the compressional waves. Despite the apparent utility of shear-wave seismic data for seismic stratigraphic analyses, a variety of logistical and cost reasons prevent them from being more widely em-

Because of seismic resolution limits and the ambiguity associated with seismic facies (i.e., any one seismic facies can be associated with several types of stratigraphic deposits), seismic stratigraphic analyses should be integrated with wireline logs, core, and other data types wherever possible. Figure 7 shows an example of this approach. In this Cretaceous clastics example, log-based correlations were ambiguous because of the highly channelized nature of the deposits. Conversely, some important stratigraphic surfaces could not be defined in the seismic data because of resolution issues and, perhaps, imaging problems in this P-wave data set. Integration of the log and seismic data, and a modern analog, reduced the ambiguity in the seismic stratigraphic correlations and led to more confident lithology predictions in interwell areas. These geometry-based approaches tend to be most commonly applied by interpreters with geologic backgrounds. Vertical transects (e.g., 2D seismic data) continue to be used in exploration settings to help define lithologies and deposition al histories in areas lacking well control (e.g., Bachtel et al., 2004; Gregersen and

w o D

ployed (e.g., Harris and O ’Brien, 2008 ).

Skaarup, 2007). Diagnostic combinations of seismic facies and reflection terminations have been advanced for a variety of different depositional settings (e.g., Handford and Loucks, 1993 ; Weimer and Slatt, 2004; Figure 8). However, the approach has some pitfalls, including: (1) nonuniqueness of the seismic response due to resolution problems, interference effects

4 Converted-wave seismic data generally involve using a compressional-wave source to generate downgoing seismic energy but recording upgoing shear waves that were produced by mode conversion at stratigraphic boundaries. See Margrave et al. (1998) for more details.


/ rg o . g es . y r ar b l/i :/ p tt h esat U f o s rm e T ee s ;t h g ir y p o c r o es en ilc G E S o t ct je

or other phenomena (Figure 9); (2) nonuniqueness of the genetic relationships between depositiona l process, external form/seismic facies and lithology, (3) nonuniqueness of the relationship between depositional facies (e.g., turbidite sheets) and physical properties of interest (e.g., porosity and permeability are affected by the degree/type of diagenesis, which is not readily predicted from depositional systems approaches), and (4) nonquantitative output. Finally, although some slices/attribu te extractions show geomorphic features that are readily identifiable as depo-

sitional features, in some cases, these views show ambiguous patterns that can be interpreted in several different ways.

Seismic attributes and feature detection Not all stratigraphic features of interest are readily apparent in amplitude data. Various seismic attributes have shown to be useful for detecting stratigraph ic features in the data in the same way that some attributes are useful for identifying structural features such as faults. Unfortunately, attribute-based analyses are not

u sb n o ti u ib rt si d e R . 1 6 . 8 9 . 4 6 .1 8 9 o t 4 1 / 8 1 / 7 0 d e ad lo n w o D Figure 7. Integration of a vertical transect, stratal slice, wireline logs, and suspected modern analog to define meandering fluvial

point bar deposits. (a) Seismic transect showing key seismic stratigraphic surfaces mapped through the integration of seismic and log data. (b) The shingled reflections in the yellow oval correspond to shalier-upward successions in gamma-ray logs (c). A slice through the data at this level (d) shows crescentic amplitude patterns that are similar in scale and planform morphology to scroll bars of a modern meandering river system (e). See Sarzalejo and Hart (2006) for fuller discussion. Interpretation / August 2013 SA9


typically used by geology-based interpreters (i.e., see papers published in the sedimentary geology or petroleum geology literature). I explore the application of seismic attributes to seismic stratigraphy in this section. I defer discussion of inversion results to a later section, although these derived volumes are sometimes referred to as “physically significant” attributes. Several definitions of seismic attributes exist, but gener ally they can be thought of as quantitative

Fi gu re 8. Seismic lapout geometries and

facies common to carbonate platform deposits. Labeled features are: (1) karst-related truncations, (2) shelf mounds, (3) landward migrating clinoforms (rimmed shelves), (4) bioherms (rimmed shelves), (5) steep depositional slopes ( >angle of repose), (6) downlapping clinoforms at toe-of-slope, (7) alternating downlap/onlap, (8) convergence of clinoform reflections, (9) shelf edge incision, and (10) incision within sequences. Redrawn from Handford and Loucks (1993). Reprinted by permission of the AAPG, whose permission is required for further use.

Figure 9. As demonstrated in this example,

geometry-based seismic stratigraphic interpretations can be ambiguous because of seismic resolution problems, interference effects, or other factors. Top row shows a simple geologic model composed of layers having different acoustic properties. Image at top right shows how the stratigraphy can be subdivided into three packages: (1) a lowermost unit consisting of folded/dipping layers (yellow), (2) a middle unit consisting of a divergent basin fill (green), and (3) an undisturbed upper unit consisting of horizontally layered strata (blue). An unconformity (dashed red line) separates the uppermost unit from the underlying two units. The middle row shows the stratigraphy as imaged using a 75 Hz Ricker wavelet. The reflections in this model generally show the structural/stratigraphic geometries of the model, and the image allows the three stratigraphic units to be identified (middle right). However, reflections appear to converge in the middle unit (might be interpreted as stratigraphic pinchouts?) and there appears to be some subtle relief associated with the unconformity (might be interpreted as differential erosion of harder/softer layers?). The Lowermost row shows the stratigraphy as imaged using a 25 Hz Ricker wavelet. In this image, it would be possible to interpret the presence of a sequence-bounding unconformity (red dashed line, lower right) that has truncation below and onlap above (reflection polarity would need to be considered). Only two stratigraphic units might be inferred from this interpretation. Based on seismic modeling presented in Hart (2000).


measures derived from the seismic data or interpretations. They include simple amplitude extractions (e.g., Figure 3, complex-trace attributes, other mathematical manipulations of the seismic trace (e.g., integration and derivative of the seismic trace; Figure 10), and measures derived from interpretations (e.g., curvature; Figure 11). Summaries and examples of seismic attributes are presented by Brown (2004) , Chopra and Marfurt (2007), Hart (2011), and others. Meta-attributes are


Fig ure 10. Poststack attributes used for detection of stratigraphic features based on simple seismic modeling. (a) Geologic model

of a prograding succession of clinoforms. High-velocity sands (yellow) grade down into low-velocity shale (brown and gray). (b) Predicted amplitude response (variable density display with wiggle overlay) of the clinoforms using a 60 Hz Ricker wavelet. (c) Integrated trace (relative acoustic impedance) attribute derived from the amplitude data shown in (b). The sands correspond to relatively high impedance (green) and the shales to low impedance (red). Wiggle trace overlay shows the srcinal seismic amplitudes. The clinoform geometry is apparent. (d) Predicted amplitude response (variable density display with wiggle overlay) of the clinoforms using a 30 Hz Ricker wavelet. Subtle changes in amplitude are produced by the changes in sand thickness, but the clinoform geometry is not clearly visible. (e) Integrated trace (relative acoustic impedance) attribute derived from the amplitude data shown in (d). The upper sandy portion of the succession corresponds to relatively high impedance (green) and the shales to low impedance (red), but the clinoform geometry is not apparent. Wiggle trace overlay shows the srcinal seismic amplitudes. (f) Second derivative attribute derived from the amplitude data shown in (d). This attribute is capturing subtle changes in waveform that suggest the presence of the clinoforms. From Hart (2008b). Interpretation / August 2013 SA11


Figur e 11.Example of horizon curvature for stratigraphic

analysis. (a) Time-structure map of the top of a leveed-channel complex. (b) Subtle structural and stratigraphic features are emphasized when dip curvature is overlain over the surface and directional lighting is applied. (c) Expanded and rotated view of the dip curvature overlay showing some fine-scale morphological features that are emphasized by curvature visualization. Note the changes in sharpness of the channel margin (green and blue) along its length, and the presence of an incised inner channel (shown by red hues indicating negative curvature) in the upper portion of this image. Reproduced with permission from Hart and Sagan (2007).


attributes derived by the combination of more than one attribute ( de Rooij and Tingdahl, 2002 ; Figure 12). Qualitative analyses of reflection amplitude and frequency, in addition to geometric analyses, originally formed part of seismic facies analyses ( Mitchum et al., 1977). Taner and Sheriff (1977) are generally credited with introducing the seismic stratigrap hy community to complex trace attributes. Other workers subsequently incorporated complex trace attributes into seismic stratigraph ic analyses (e.g., Fontaine et al., 1987) but published examples are few. Similarly, relatively few attempts have been made to define seismic attributes that capture reflection geometries such as parallelism, convergence, or downlap (e.g., Barnes, 2000; van Hoek et al., 2010 ). The seismic response is frequency-dependent. Although some analyses have examined changes in reflection configuration in vertical transects as a function of frequency (e.g., Zeng, 2013), most published analyses have focused on map view slices through 3D volumes. Spectral decomposition methods break the seismic signal down into its component frequencies. The resultant images commonly bring out stratigraphic (or structural) features that are not readily appar ent in the srcin al broadband data (Castagna and Sun, 2006 ; Chopra and Marfurt, 2007 ). Different approaches are possible for analyzing the results, but one common approach is to optically stack different frequency bands (e.g., Figure 13). Frequency-dependent tuning observed in these analyses can be used to predict the thickness of stratigraphic features. Automated seismic facies analyses use a variety of computer-based techniques to characterize seismic trace shape. Thereafter, it is assumed (or hoped) that each facies can be related to lateral variations in lithology, rock properties, and/or fluid content of the stratigraphic bodies being imaged. Several different approaches have been applied to this task, including artificial intelligence-based methods such as neural networks (Coléou et al., 2003; Figure 14). Originally applied to a narrow time window defined with respect to a single reflection, 3D applications of seismic facies analysis also have been developed (e.g., Farzadi and Hesthammer, 2007; Gao, 2007 ). These automated seismic facies can be very useful; however, knowledge of depositional systems is required to: (1) help define geologically meaningful windows for the facies analyses, and (2) ensure that the stratigraphic features interpreted to be revealed by the analyses make geologic sense. References presented in this section about seismic attributes are generally from the geophysical literature. Despite the potential applications to sedimentary geol-a ogy, seismic attrib ute studie s have yet to become mainstream part of that discipline ’s toolkit.

Physical properties prediction While indirect methods for predicting physical properties from seismic data (i.e., exploiting relationships


between seismic facies, depositional processes, and inversion result from the Western Canada Sedimentary lithology) were being developed and exploited by Basin where Cretaceous siliciclastic deposits (fluvial, sedimentary geologists and geophysicists, other geoestuarine, and shallow-marine deposits) unconformphysicists were developing ways of making direct, ably overlie a succession of Devonian carbonates. quantitative predictions of rock properties from seismic The contact between those two packages (black bar data. Seismic inversion methods are used to transform seismic data into quantitative predictions of rock propertie s (e.g., acoustic impedance, lithology, porosity, fluid saturation). Several different approaches are available for this purpose, with one of the principal distinctions being whether the inversion is undertaken before or after the seismic data have been stacked. Prestack methods are used to define elastic properties of the strata being imaged. Poststack methods are generally used to predict acoustic impedance. Most inversion methods are deterministic; however, stochastic inversion methods (e.g., incorporati ng elements of geostatistics; e.g., Rowbotham et al., 2003 ) also have been developed. Rock physics analyses and Figure 12.Sweetness attribute applied to a deepwater clastics data set. Sweetness is a meta-attribute derived by combining instantaneous amplitude and inmodeling are needed components of instantaneous frequency. In this example, modified from Hart (2008a), the image at version studies. Selected summaries of left (a) shows an inclined slice through a submarine fan system as seen in the seismic inversion methods include those srcinal amplitude display. The image at right (b) shows an identical slice of Veeken and DaSilva (2004) , Sen through a sweetness attribute volume, corendered with semblance to highlight (2006), and Bosch et al. (2010) . channel margins. High sweetness values (yellow/orange) are associated with There are several interrelated reasands of at least two crosscutting channel systems. These channe ls are not sons for inverting seismic data to acousreadily apparent in the srcinal amplitude volume (a). Reprinted by permission of the AAPG, whose permission is required for further use. tic impedance data prior to commencing


stratigraphic analyses (Latimer al., 2000). Acoustic impedance is a etrock property that can typically be directly related to lithology, porosity, and pore-filling fluids. Second, seismic data that have been inverted to acoustic impedance show layer properties rather than interface properties. Reflection s in conventional seismic data correspond to interfaces where there are changes in acoustic impedance. The reflections do not image the layers themselves. After inversion, the seismic image should show the acoustic properties of the layers themselves. As such, the data more closely resemble a geologic cross section. Finally, most inversion methods incorporate methods to remove the “blurring” effect of the seismic wavelet,

w o D

thereby improving seismic resolution and minimizing tuning effects. Thin beds, not visible in the srcinal seismic data, can become apparent in the inverted seismic data. Figure 15 compares a seismic image with a model-based acoustic-impedance

Figure 13. Spectral decomposition example from Moser et al. (2004) . Ampli-

tude variations for three different frequencies are displayed in red, green, and blue then overlain. The resultant image shows stratigraphic features that are not readily visible in the srcinal broadband seismic data. No scale bar was provided by the srcinal authors, but it seems reasonable to assume that the width of the image is probably 3–5 km across (meandering channels, such as those seen in the image, are typically several tens of meters wide). Interpretation / August 2013 SA13


near base of each well display) is not readily apparent in the srcinal seismic data. The unconformity becomes readily apparent after the inversion. Clearly the impedance version of the data would be much more useful for seismic stratigraphic analyses than the srcinal amplitude volume. Prestack inversion methods solve for P-impedance, S-impedance, and density. These products can then be manipulated in various ways to predict elastic properties (Young ’s modulus, Poisson ’s ratio, VP/VS ratio, etc.) that can be related to lithology, porosity, fluid saturation, etc. Given the current interest in unconventional reservoirs (e.g., “shale plays ”), there is much interest in using inversion products to predict areas

nonuniqueness of the inversion solution, and (4) nonuniqueness of the relationship between elastic/acoustic properties and lithologic properties of interest to seismic stratigraphers. Geophysicists have fully embraced

of better reservoir properties (porosity, saturation) and to design hydraulic fracturehydrocarbon treatments (e.g., Close et al., 2012 ; Figure 16). Although seismic inversion products have become widely used in the petroleum industry, there are several ambiguities and limitations that need to be kept in mind. These problems include: (1) dependence of the result on the low-frequency model, (2) wavelet estimation, (3)

Fig ure 15. Example showing how a simple model-based

acoustic impedance inversion was used to help with a stratigraphic interpretation (i.e., mapping a significant unconformity). (a) The srcinal seismic amplit ude data. Well control indicates the presence of a significant unconformity near the base of the wells that separates Cretaceous clastics from underlying Devonian carbonates (black line and star). The unconformity does not correspond to a prominent reflection in the data. (b) Model-based inversion result showing the


same presented in (a).between The unconformity is clearly visible asprofile an upward transition high-impedance carbonates (purple) and the overlying clastics (blue, red, yellow, green). A high-impedance layer (purple) above the unconformity corresponds to carbonate-cemented sandstone. Wireline logs in both cases are sonic logs. From Hart (2011) . Reprinted by permission of the AAPG, whose permission is required for further use.

w o D Figur e 14. Sample automated seismic facies classification

results from Marroquin et al. (2009). (a) Devonian pinnacle reefs and (b) a probable Jurassic tidal channel (siliciclastic). Although different seismic facies (and lithologies) are present at the two stratigraphic levels, the software has used the same color palette for both classification exercises.


Figure 16. Prestack elastic inversion-based prediction of the

distribution of “brittle” and “ductile” rocks in the Cretaceous Eagle Ford Formation. Map area covers approximately 2 80 km . Modified from Treadgold et al. (2011).

inversion methods for predicting rock properties (including fluid content) but the sedimentary geology community has been slow to adopt these methods for studying depositional systems (but see Contreras and Latimer,


2010). Sedimentary geologists could play a key role in inversion studies by helping to define which results are geologically/stratigraphically possible. Another approach to physical property prediction was termed the “data-driven” methodology by Schultz et al. (1994). This approach exploits empirically derived relationships between seismic attributes and log-derived properties of interest to predict those properties away from well contro l (e.g., Hampson et al., 2001 ). Tebo and Hart (2005) and Sagan and Hart (2006) demonstrate how this approach could be used to study porosity development in two different carbonate settings, and Sarzalejo Silva and Hart (2013) show how the approach could be used to evaluate a siliciclastic heavy oil reservoir. Although once relatively widely used, the popularity of this empiric al approach has waned as geophysicists ’ attention has turned to focus on rock physics-based inversion approaches described previously. The sedimentary geology community did not embrace attribute-based property predictions.

Stratigraphic controls on seismic response Despite much overlap in topics of interest, there has been relatively little collaboration between the sedimentary geology community, rock physicists, and geophysicists in terms of understanding the relationships between depositional/diagenetic processes and seismic responses. This, despite the fact that sediment provenance (i.e., the source of the sediment), depositional processes, and diagenesis are the only controls on mineralogy, porosity, fabric, and other properties that, in turn, are fundamental controls on density, elastic moduli, anisotropy, etc. Figure 17. Two stacked “coarsening upward successions ”

(parasequences) exposed in Cretaceous clastics of the Book Cliffs, Utah. Each succession is shaley at the base with sandstone beds generally becoming thicker and more abundant upward. People at lower right for scale. Note that these parasequences would be below the resolution of most petroleum industry seismic data sets.

Rock physicists have developed many different mathematical treatments for predicting properties of interests based on theoretical mixes of grain sizes, porosity, mineralogy, etc., in siliciclastic successions (e.g., Mavko et al., 1998 ; Avseth et al., 2005 ). However, there are few studies that specifically link depositional processes (e.g., debris flows, tractive sediment transport) or Figure 18.Outcrop photo of the Eagle Ford

Formation, west of Del Rio, Texas, from Hart et al. (this issue). The formation is not a lithologically homogeneous unit. Instead, different lithologies of different physical properties (porosity, Poisson’s ratio, Young’s modulus, etc.) are arranged into laminae, beds, bedsets (shown), and members, all of which are shorter than a seismic wavelet. Do maps of seismically derived physical properties, such as the one shown in Figure 16, represent average properties over the entire interval or are they the product of some other phenomenon, such as changes in stratigraphic stacking patterns that affect the AVO response (e.g., Figure 20) upon which the physical property predictions are based?


diagenetic processes (e.g., effects of chlorite rims on porosity development) to an appropriate physical model (but see, for example, Avseth et al. [2005] , p. 83–90). Eberli et al. (2003) and Weger et al. (2009) examined relationships between carbonate porosity / rg o . g es . ry ar b l/i :/ p tt h esat U f o s rm e T ee s ;t h g ir y p o c r o es en ilc G E S o t ct je

Figur e 21. Example of how seismic geometries can help

guide log correlations. In this simple case, gamma ray logs from a series of wells (top row) can be correlated in at least two different ways (middle row). A seismic transect from this area (bottom) shows clinoforms that would support the “shingled” correlation option of the middle row. Modified from Hart (2011) and reprinted by permission of the AAPG, whose permission is required for further use.

Fig ure 19. Simple 1D acoustic models showing the predicted

seismic response of various stratigraphic stacking patterns (blocky, fining upward, coarsening upward) as a function of thickness. Each block shows an acoustic impedance profile (increasing to right) and the calculated seismic response using a simple Ricker wavelet. From Hart (2008b).

types (a function of depositional and diagenetic processes) and rock physical properties. Closer integration between sedimentary geologists and rock physicists might allow more realistic modeling of mineral textures and would almost certainly help to ensure that appropriate rock physics models are selected for property prediction in seismic modeling studies, i.e., certain mineral textures (e.g., clay-supported sandstones) are more likely to develop in some depositional facies than others. For example, Hart et al. (this issue) state that mathematical/conceptual models developed for clay-rich shales are inappropriate for source-rock reservoirs (shale plays) because the latter are commonly clay poor. In fine-grained systems


(shales), there has been considerable work undertaken to examine the origins and effects of anisotropy at the particle scale (e.g., Sayers, 2005 ; Day-Stirrat et al., 2010). Layering imparts a vertical axis of symmetry, making sedimentary rocks vertically transverse isotropic media5. The anisotropic parameters are derived from measurements on core plugs and need to be upscaled mathematically (e.g., Backus, 1962 ; Berryman, 2008) to predict seismic-scale properties that cannot readily be measured otherwise. Sedimentary geologists could help to define realistic stratigrap hic model parameters for this type of upscaling work. Sedimentary rocks are typically bedded, with collections of beds forming bedsets or parasequences (Figures 17 and 18). Seismic modeling has been used successfully to predict seismic responses (amplitude and other attributes) for various types of stratigraphic successions (e.g., Meckel and Nath, 1977 ; Hart and Chen, 2004; Figure 19), typically with a focus on predict-

w o D

ing poststack character. However, variations in stratigraphic layering also are known to affect amplitude variation with offset responses that would be visible in prestack analyses (Figure 20). The effects of stratigraphic

Figur e 20. Amplitude-variation-with-offset (AVO) responses

for various stratigraphic stacking patterns. Although the same change in physical properties is present in each case, the AVO responses are markedly different. Modified from Lindsay and Van Koughnet (2001).


5 I.e., properties such as lithology, porosity, Poisson’s ratio, and others change more rapidly from bed to bed, rather than along beds.

variability on AVO responses, or physical property predictions (e.g., elastic inversion) derived from AVO analyses, are seldom documented explicitly. •

Conclusions / rg o . g es . ry ar b l/i :/ p tt h esat U f o s rm e T ee s ;t h g ir y p o c r o es en ilc G E S o t ct je u sb n o ti u ib rt si d e R . 1 6 . 8 9 . 4 6 .1 8 9 o t 4 1 / 8 1 / 7 0 d e ad lo n w o D

The field of seismic stratigraphy is a mature science that has proven to be useful in a variety of exploration, development, and fundamental studies in sedimentary geology. The four main elements of seismic stratigraphy are seismic sequence analysis, seismic facies analysis, reflection character analysis, and seismic geomorphol ogy. The first three were developed for use on 2D seismic data and can be transferred to 3D volumes for use on vertical transects. Seismic geomorphology evolved from map-v iew analyses of 3D seismic volume s and has since expanded to incorporate a variety of visualization and analysis techniques. Unfortunately, a variety of factors can lead to nonunique interpretations, and physical property predictions are typically qualitative or, at best, probabilistic in nature. Perhaps for these reasons, the physics-based quantitative output of seismic inversion methods has become favored for rock property prediction in the petroleum industry. On the other hand, seismic data availability and quality problems preclude the application of inversion methods in all settings. The field of seismic stratigraphy, as practiced by sedimentary geologists, could benefit from more routine integration of seismic attribute studies and seismic-based physical property predictions. To do so, sedimentary geologists will need to learn the physics behind these methods. The sedimentary geology community has embraced physics before, such as applications in sediment transport and bedform development (e.g., Middleton and Southard, 1984 ) or basin analysis (e.g., Allen and Allen, 2005 ), and perhaps will again. I have tried to highlight a variety of geoscience subdisciplines where the methods and interests of sedimentary geologists and geophysicists overlap. Here, I present several reasons why enhanced collaboration between these two groups could be mutually advantageous. •

Enhanced fundamental understanding of depositional systems. As described above, the sedimentary geology community has been very slow to use/accept seismic analyses that are based on (qualitative) attribute studies or (quantitative) physical property predictions. This is unfortunate because these analyses can provide clear images of stratigraphic features and quantitative measures that would be very useful for fundamental studies of depositional systems. Miall (2002) , for example, derived a variety of quantitative morphologic parameters from meandering fluvial systems imaged in time slices through a 3D seismic amplitude volume. Wood and Mize-Spansky (2009) did likewise for a deepwater leveedchannel system. Enhanced seismic stratigraphic

•

•

•

analyses, for example using volume visualization on attribute volumes, might provide new information about other depositional systems that cannot be adequately imaged using amplitude displays. Better predictive capabilities in exploration settings. The high-quality 3D seismic data needed to apply prestack elastic inversion methods are not available everywhere. The results of any given inversion study can be used to provide analog data for risk assessment in other areas but should be employed only if appropriate analogs can be identified. For example, it would be inappropriate to use data from a sand-prone submarine fan system to assess risk in a mud-prone system. Sedimentary geologists play an important role in selecting appropriate analogs. Improved geology-based interpretations. Geologic interpretations based on subsurface data (wireline logs, core) have an inherent degree of ambiguity because of data availability (or lack thereof) and, typically, data that can be interpreted in more than one way. Seismic data can provide interwell information that can be critical for assembling a meaningful geologic story (e.g., Figure 21). For example, log-based interpretations are problematic for some deep “shale basins” that are essentially undrilled. Seismic data also can provide useful analog dimensional and morphology data for features that cannot be accurately mapped with available subsurface control. Improved geophysics-based interpretations. Geophysics-based rock property predictions (e.g., inversion) are likewise associated with ambiguity because of reasons such as data quality, data availability, assumptions made during the inversion (e.g., wavelet estimation), etc. Stratigraphic analyses of seismic-based rock property predictions should always be undertaken to ensure that those predictions are geologically reasonable. For example, it would be appropriate to ask whether the predicted distribution of brittle and ductile rocks shown in Figure 16 makes geologic sense, given what is known about Eagle Ford stratigraphy (e.g., Figure 18) and depositional processes. Subseismic prediction of rock properties . Many reservoirs contain internal baffles, thief zones, variations in porosity, or other stratigraphic heterogeneities that can have important economic impacts but are below seismic resolution (e.g., Figures 10, 17, and 18). Seismic data can provide the structural and stratigraphic context in these cases, but stratigraphic knowledge plays an important in predicting and distributionrole of these features. the presence

References Shell Oil Company, 1987, Atlas of seismic stratigraphy, in Bally, A. W. ed., AAPG Studies in Geology 27, 15 –71. Interpretation / August 2013 SA17

Allen, P. A., and J. R. Allen, 2005, Basin analysis: Principles and application: Blackwell. Avseth, P., T. Mukerji, and G. Mavko, 2005, Quantitative seismic interpretation: Cambridge University Press. Bachtel, S. L., R. D. Kissling, D. Martono, S. P. Rahardjanto, P. A. Dunn, and B. A. MacDonald, 2004, Seismic stratigraphic evolution of the Miocene –Pliocene Segitiga Platform, East Natuna Sea, Indonesia: The origin, growth, and demise of an isolated carbonate platform, in G. P. Eberli, J. L. Masaferro, and J. R. Sarg, eds., Seismic imaging of carbonate reservoirs and systems: AAPG Memoir 81, 309 –328. Backus, G. E., 1962, Long-wave elastic anisotropy produced by horizontal layering: Journal of Geophysical

of fluvial reservoirs: AAPG Bulletin, 96, 2143 –2166, doi: 10.1306/04211211179. Contreras, A. J., and R. B. Latimer, 2010, Acoustic impedance as a sequence stratigraphic tool in structurally complex deepwater settings: The Leading Edge, 29, 1072–1082, doi: 10.1190/1.3485768. Cross, T. A., and M. A. Lessenger, 1988, Seismic stratigraphy: Annual Review of Earth and Planetary Sciences, 16, 319 –354, doi: 10.1146/annurev.ea.16.050188.001535. Day-Stirrat, R. J., S. P. Dutton, K. L. Milliken, R. G. Loucks, A. C. Aplin, S. Hillier, and B. A. van der Pluijm, 2010, Fabric anisotropy induced by primary depositional variations in the silt: Clay ratio in two fine-grained slope fan complexes: Texas Gulf Coast and northern North Sea:

Research, 67, 4427 –4440, doi: 10.1029/JZ067i011p04427. Barnes, A. E., 2000, Attributes for automating seismic facies analysis: 70th Annual International Meeting, SEG, Expanded Abstracts, 553 –556. Berryman, J. G., 2008, Exact seismic velocities for transversely isotropic media and extended Thomsen formulas for stronger anisotropies: Geophysics, 73, no. 1, D1 –D10, doi: 10.1190/1.2813433. Bertram, G. T., and N. J. Milton, 1996, Seismic stratigraphy, in D. Emery, and K. J. Meyers, eds., Sequence stratigraphy: Blackwell, 45–60. Bosch, M., T. Mukerji, and E. F. Gonzalez, 2010, Seismic inversion for reservoir properties combining statistical rock physics and geostatistics: A review: Geophysics, 75, no. 5, 7 5A165 –75A176, doi: 10.1190/1 .3478209. Brown, A. R., 2004, Interpretation of 3-D seismic data, 6th ed.: AAPG Memoir 42.

Sedimentary Geology, 226, 42 –53, doi: 10.1016/j.sedgeo .2010.02.007. de Rooij, M., and K. Tingdahl, 2002, Meta-attributes — the key to multivolume, multiattribute interpretation: The Leading Edge, 21, 1050 –1053, doi: 10.1190/1.1518445. Eberli, G. P., G. T. Baechle, F. S. Anselmetti, and M. L. Incze, 2003, Factors controlling elastic properties in carbonate sediments and rocks: The Leading Edge, 22, 654 –660, doi: 10.1190/1.1599691. Farzadi, P., and J. Hesthammer, 2007, Diagnosis of the Upper Cretaceous palaeokar st and turbidite systems from the Iranian Persian Gulf using volume-based multiple seismic attribute analysis and pattern recognition: Petroleum Geoscience, 13, 227 –240, doi: 10.1144/ 1354-079306-710. Fontaine, J. M., R. Cussey, J. Lacaze, R. Lanaud, and L. Yapaudjian, 1987, Seismic interpretation of carbonate depositional environments: AAPG Bulletin, 71,


Brown, A. R., C. G. Dahm, and R. J. Graebner, 1981, A stratigraphic case history using three-dimensional seismic data in the Gulf of Thailand: Geophysical Prospecting, 29, 327 –349, doi: 10.1111/j.1365-2478.1981 .tb01017.x. Castagna, J. P., and S. Sun, 2006, Comparison of spectral decomposition methods: First Break, 24, 75–79. Catuneanu, O., 2006, Principles of sequence stratigraphy: Elsevier. Chopra, S., and K. J. Marfurt, 2007, Seismic attributes for prospect identification and reservoir characterization: SEG Geophysical Development Series 11. Close, D., M. Perez, B. Goodway, and G. Purdue, 2012, Integrated workflows for shale gas and case study results for the Horn River Basin, British Columbi a, Canada: The Leading Edge, 31, 556 –569, doi: 10.1190/ tle31050556.1. Coléou, T., M. Poupon, and K. Azbel, 2003, Unsupervised

281–297. Gao, D., 2007, Applicati on of three-dimensional seismic texture analysis with special reference to deep-marine facies discrimination and interpretation: Offshore Angola, west Africa: AAPG Bulletin, 91, 1665 –1683, doi: 10.1306/08020706101. Gregersen, U., and N. Skaarup, 2007, A mid-Cretaceous prograding sedimentary complex in the Sisimiut Basin, offshore west Greenland — Stratigraphy and hydrocarbon potential: Marine and Petroleum Geology, 24, 15– 28, doi: 10.1016/j.marpetgeo.2006.10.005. Gregory, A. R., 1977, Aspects of rock physics from laboratory and log data that are important to seismic interpretation: in C. E. Payton, ed., Seismic stratigraphy — application to hydrocarbon exploration: AAPG Memoir 26, 15 –46. Hampson, B., J. Schuelke, and J. Quirein, 2001, Use of multi-attribute transforms to predict log properties from

w o D

seismic facies classification: A review and comparison of techniques and implementation: The Leading Edge, 22, 942 –953, doi: 10.1190/1.1623635. Colombera, L., F. Felletti, N. P. Mountney, and W. D. McCaffrey, 2012, A database approach for constraining stochastic simulations of the sedimentary heterogeneity

seismic data: Geophysics, 66, 220 –236, doi: 10.1190/1 .1444899. Handford, C. R., and R. G. Loucks, 1993, Carbonate depositional sequences and systems tracts — responses of carbonate platforms to relative sea-level changes, in R. G. Loucks, and J. F. Sarg, eds., Carbonate sequence



stratigraphy, recent developments and applications, AAPG Memoir 57, 3–41. Hardage, B. A., and I. J. Aluka, 2006, Elastic wavefield seismic stratigraphy: AAPG Search and Discovery Article 40183, http://www.searchanddiscovery.com/ documents/2006/06005hardage_gc/imagim/hardage.pdf, accessed 15 March 2013. Harris, R., and J. O ’Brien, 2008, Imaging tight gas sandstones in the East Texas Basin: First Break, 26, 57–68. Hart, B. S., 2000, 3-D seismic interpretation: A primer for geologists: SEPM Short Course Notes, 48. Hart, B. S., 2008a, Channel detection in 3-D seismic data using sweetness: AAPG Bulletin, 92, 733 –742. Hart, B. S., 2008b, Stratigraphically significant attributes:

Mavko, G., T. Mukerji, and J. Dvorkin, 2009, The rock physics handbook: Cambridge University Press. McCullagh, T., and B. S. Hart, 2010, Stratigraphic controls on production from a basin-centered gas system: Lower Cretaceous Cadotte Member, Deep Basin Alberta, Canada: AAPG Bulletin, 94, 293 –315, doi: 10.1306/ 08260908137. Meckel, L. D., and A. K. Nath, 1977, Geologic considerations for stratigraphic modeling and interpretation, in C. E. Payton, ed., Seismic stratigraphy — Application to hydrocarbon exploration: AAPG Memoir 26, 417–438. Miall, A. D., 2002, Architecture and sequence stratigraphy of pleistocene fluvial systems in the Malay Basin, based

The Leading Edge, 27, 320 –324, doi: 10.1190/1.2896621. Hart, B. S., 2011, Introduction to seismic interpreta tion: AAPG. Hart, B. S., and M. A. Chen, 2004, Understanding seismic attributes through forward modeling: The Leading Edge, 2 3, 834 –841, doi: 10.1190/1.1803492. Hart, B. S., J. H. S. Macquaker, and K. G. Taylor, this volume, Mudstone (shale) depositional and diagenetic processes: Implications for seismic analyses of source-rock reservoirs: Interpretation (This issue). Hart, B. S., and J. A. Sagan, 2007, Curvature for visualization of seismic geomorphology, in R. J. Davies, H. W. Posamentier, L. J. Wood, and J. A. Cartwright, eds., Seismic geomorphology: Applications to hydrocarbon exploration and production: Geological Society of London Special Publication 277, 139 –149. Hart, B. S., S. Sarzalejo, and T. McCullagh, 2007, Seismic stratigraphy and small 3-D seismic surveys: The Leading

on seismic time-slice analysis: AAPG Bulletin, 86, 1201– 1216. Miall, A. D., 2010, The geology of stratigraphic sequences, 2nd Edition, Springer. Middleton, G. V., and J. B. Southard, 1984, Mechanics of sediment movement: SEPM Short Course Notes, 3. Mitchum, R. M., P. R. Vail, Jr., and J. B. Sangree, 1977, Stratigraphic interpretation of seismic reflection patterns in depositional sequences, in C. E. Payton, ed., Seismic stratigraphy — applications to hydrocarbon exploration: AAPG Memoir 26, 117 –133. Moser, J. R., F. F. Strijbos, and J. P. Castagna, 2004, Spectral variance in comparison with conventional spectral decomposition attributes: 74th Annual International Meeting, SEG, Expanded Abstracts, 441 –444. Posamentier, H. W., 2004, Seismic geomorphology: Imaging elements of depositional systems from shelf to deep basin using 3D seismic data: Implications for explora-


Edge, 2 6, 876 –881, doi: 10.1190/1.2756867. Houseknecht, D. W., and C. J. Schenk, 2004, Sedimentology and sequence stratigraphy of the cretaceous nanushuk, seabee, and tuluvak formations exposed on umiat mountain, north-central Alaska: U.S. Geological Survey Professional Paper, 1709-B, p. 18. Jol, H. M., and C. S. Bristow, 2003, GPR in sediments: A good practice guide, in C. S. Bristow, and H. M. Jol, eds., Ground penetrating radar in sediments: Geological Society of London, Special Publication 211, 9 –27. Latimer, R. B., R. Davison, and P. van Riel, 2000, An interpreter’s guide to understanding and working with seismic-derived acoustic impedance data: The Leading Edge, 1 9, 242 –256. Lindsay, R., and R. Van Koughnet, 2001, Sequential backus averaging: Upscaling well logs to seismic wavelengths: The Leading Edge, 20, 188 –191, doi: 10.1190/1.1438908. Margrave, G. F., D. C. Lawton, and R. R. Stewart, 1998, In-

tion and development: Geological Society London Memoirs, 29, 11–24, doi: 10.1144/GSL.MEM.2004.029 .01.02. Prather, B. E., J. R. Booth, G. S. Steffens, and P. A. Craig, 1998, Classification, lithologic calibration, and stratigraphic succession of seismic facies of intraslope basins, deep-water Gulf of Mexico: AAPG Bulletin, 82, 701 –728. Rowbotham, P. S., D. Marion, P. Lamy, E. Insalaco, P. A. Swaby, and Y. Boisseau, 2003, Multidis ciplinary stochastic impedance inversion: Integrating geological understanding and capturing reservoir uncertainty: Petroleum Geoscience, 9, 287 –294, doi: 10.1144/1354079302-490. Sagan, J. A., and B. S. Hart, 2006, Three-dimensional seismic-based definition of fault-related porosity development: Trenton–Black river interval, Saybrook, Ohio: AAPG Bulletin, 90, 1763 –1785, doi: 10.1306/

w o D

terpreting channels sands with 3C-3D seismic data: The Leading Edge, 17, 509 –513, doi: 10.1190/1.1438000. Marroquin, I. D., B. S. Hart, and J. J. Brault, 2009, A visual mining-based methodology to conduct seismic facies analysis, part 2: Application to 3-D seismic data: Geophysics, 74, no. 1, P13 –P23, doi: 10.1190/1.3046456.

07190605027. Saller, A. H., J. T. Noah, A. P. Ruzuar, and R. Schneider, 2004, Linked lowstand delta to basin-floor fan deposition, offshore Indonesia: An analog for deep-water reservoir systems: AAPG Bulletin, 88, 21–46, doi: 10.1306/ 09030303003.



/ rg o . g es . ry ar b l/i :/ p tt h esat U f o s rm e T ee s ;t h g ir y p o c r o es en ilc G E S o t ct je u sb n o ti u ib rt si d e R . 1 6 . 8 9 . 4 6 .1 8 9 o t 4 1 / 8 1 / 7 0 d e ad lo n

Sarzalejo, S., and B. S. Hart, 2006, Stratigraphy and lithologic heterogeneity in the Mannville group (southeast Saskatchewan) defined by integrating 3-D seismic and log data: Bulletin of Canadian Petroleum Geology, 54, 138 –151, doi: 10.2113/gscpgbull.54.2.138. Sarzalejo Silva, S. E., and B. S. Hart, 2013, Advanced seismic-stratigraphic imaging of depositional elements in a lower cretaceous (Mannville) heavy oil reservoir, westcentral Saskatchewan, Canada, in F. J. Hein, D. Leckie, S. Larter, and J. Suter, eds., Heavy-oil and oil-sand petroleum systems in Alberta and beyond: AAPG Studies in Geology, 64, 359 –372. Sayers, C. M., 2005, Seismic anisotropy of shales: Geophysical Prospecting, 53, 667 –676, doi: 10.1111/j.1365-

Treadgold, G., B. Campbell, B. McLain, S. Sinclair, and D. Nicklin, 2011, Eagle Ford shale prospecting with 3D seismic data within a tectonic and depositional system framework: The Leading Edge, 30, 48 –53, doi: 10.1190/1 .3535432. Vail, P. R., J. Hardenbol, and R. G. Todd, 1984, Jurassic unconformities, chronostratigraphy and sea-level changes from seismic stratigraphy and biostratigraphy, in J. S. Schlee, ed., Interregi onal unconformities and hydrocarbon accumulation, AAPG Memoir 36, 129–144. van Hoek, T., S. Gesbert, and J. Pickens, 2010, Geometric attributes for seismic stratigraphic interpretation: The Leading Edge, 29, 1056 –1065, doi: 10.1190/1.3485766.

2478.2005.00495.x. Schultz, P. S., S. Ronen, M. Hattori, and C. Corbett, 1994, Seismic-guided estimation of log properties: Part 1: A data-driven interpretation methodology: The Leading Edge, 13, 305 –310, doi: 10.1190/1.1437020. Sen, M., 2006, Seismic inversion: SPE. Snedden, J. W., and J. F. Sarg, 2008, Seismic stratigraphy — a primer on methodology: AAPG Search and Discovery, 40270, accessed 8 July 2008. Taner, M. T., and R. E. Sheriff, 1977, Application of amplitude, frequency and other attributes to stratigraphic and hydrocarbon determination, in C. E. Payton, ed., Seismic stratigraphy: Applications to hydrocarbon exploration: AAPG Memoir 26, 391 –327. Tebo, J. M., and B. S. Hart, 2005, Use of volume-based 3-D seismic attribute analysis to characterize physical property distribution: A case study to delineate reservoir heterogeneity at the Appleton field, SW Alabama: Jour-

Veeken, P. C. H., and M. DaSilva, 2004, Seismic inversion and some of their constraints: First Break, 22, 47–70. Weger, R. J., G. P. Eberli, G. T. Baechle, J. L. Massaferro, and Y.-F. Sun, 2009, Quantification of pore structure and its effect on sonic velocity and permeability in carbonates: AAPG Bulletin, 93, 1297 –1317, doi: 10.1306/ 05270909001. Weimer, P., and R. M. Slatt, 2004, Petroleum systems of deepwater settings: SEG Distinguished Instructor Series no. 7. Wood, L. J., and K. L. Mize-Spansky, 2009, Quantitative seismic geomorphology of a quaternary leveed-channel system, offshore eastern Trinidad and Tobago, northeastern South America: AAPG Bulletin, 93, 101 –125, doi: 10.1306/08140807094. Zeng, H., 2013, Frequency-dependent seismic-stratigraphic and facies interpretation: AAPG Bulletin, 97, 201 –221, doi: 10.1306/06011212029.

nal of Sedimentary Research, 75, 723 –735, doi: 10.2110/ jsr.2005.058. Tipper, J. C., 1993, Do seismic reflections necessarily have chronostratigraphic significance?: Geological Magazine, 130, 47–55, doi: 10.1017/S0016756800023712.

Zeng, H., and C. Kerans, 2003, Seismic frequency control on carbonate seismic stratigraphy: A case study of the Kingdom Abo sequence, west Texas: AAPG Bulletin, 87, 273 –293, doi: 10.1306/08270201023.

w o D


SeismicStatGraph.pdf

Recommend Documents