Fractal nature of stratigraphic sequences

Wolfgang Schlager* Earth and Life Sciences, Vrije Universiteit, De Boelelaan 1085, 1081 HV Amsterdam, Netherlands

ABSTRACT distinct sequence architecture to all orders of

Orders of stratigraphic sequences are being used loosely and with widely varying def- sequences were proposed by Vail et al. (1991)
initions. The orders seem to be subdivisions of convenience rather than an indication of and Duval et al. (1998).
natural structure. It is proposed that, at least at time scales of 103–106 yr, sequences and Standard sequences, parasequences, and
systems tracts are scale-invariant fractal features in which units bounded by exposure simple sequences were first defined on silici-
surfaces and units bounded by flooding surfaces are about equally likely. Support for this clastic data sets. However, carbonates have
conceptual model derives from the well-known facts that (1) sediment architecture is large- several properties that make them particularly
ly scale invariant over a wide range of scales in time and space; (2) first-order trends of useful for the study of exposure and flooding
sea-level movements and sedimentation rates are fractal on all geologically relevant time events. (1) Carbonates are highly sensitive to
scales; and (3) detailed studies in carbonates indicate approximately equal abundance of exposure because above sea level, carbonate
flooding surfaces and demonstrable exposure surfaces among unit boundaries. Finally, production rapidly goes to zero. (2) Marine
traces of prograding shelf edges that step up and step down in response to sea-level chang- deposits are rapidly lithified upon exposure,
es are demonstrated in this study to have a fractal nature. and the cements of this terrestrial lithification
are vastly different from the marine cements
Keywords: fractals, sedimentation rates, sedimentology, sequence stratigraphy. encountered on a marine transgressive surface.
(3) Terrestrial erosion proceeds largely by dis-
INTRODUCTION even larger. Moreover, the values do not seem solution, which creates much internal porosity
When introducing the principles of se- to converge with time and improving data. but retards destruction of the preceding high-
quence stratigraphy, Vail et al. (1977) noted Figure 2 plots the durations of sequence cy- stand surface.
that sequences may vary in thickness from cles of 2nd and 3rd order on the sea-level Figure 3 presents data from carbonate
1023 to 103 m and in duration from ,106 yr to curve of Haq et al. (1987). The two categories rocks, showing the proportions of two types
108 yr. They proposed a hierarchy of orders of clearly differ in their modes but broadly over- of boundaries for sequences in the range of
different duration to subdivide the sequence- lap in range. Hardenbol et al. (1998) aban- ;104–106 yr.
stratigraphic record. Since then, orders defined doned the subdivision into 2nd- and 3rd-order Parasequence boundary. The depositional
by duration have been widely used. In addi- cycles, rightly pointing out that better under- environment shoals to shallow-marine or
tion, it was proposed that sequences of differ- standing of mechanisms is required to justify supratidal conditions; after some time of non-
ent orders also differ in depositional architec- this classification. deposition, the area is reflooded, and marine
ture (Van Wagoner et al., 1990; Vail et al., Soon after their introduction as categories deposition is resumed.
1991). in time, sequence orders were also character- Standard sequence boundary. Marine
This essay presents a critique of the concept ized by their depositional architecture. An im- sedimentation is terminated by relative sea-
of orders in sequence stratigraphy and argues portant step was the recognition of parasequ- level fall and exposure (forced regression).
that the succession of sequences is fractal rath- ences—building blocks of classical sequences After a hiatus with terrestrial conditions, the
er than a hierarchy of orders. The argument that were made up of shoaling successions area is reflooded, and marine sedimentation is
rests on four components: (1) The duration of bounded by flooding surfaces (Van Wagoner resumed. Soil features such as root casts, cal-
the presumed orders varies widely, even with- et al., 1990). However, it was also observed crete crusts, and low d13C values indicating
in one publication. (2) Exposure surfaces and that in some instances, the building blocks of influx of soil carbon are considered diagnostic
flooding surfaces as unit boundaries are both 3rd-order sequences were ‘‘simple sequences’’ of terrestrial exposure. Desiccation cracks are
common in a wide range of temporal scales. bounded by exposure surfaces (Vail et al.,
(3) Extensive studies on sea-level fluctuations 1991, p. 630). Classifications that assigned
and sedimentation rates have shown that the
principal trends of both are fractal. (4) Limited
data on shelf edges that prograde and step up
and down in response to sea level indicate that
these traces, too, are fractal.

ORDERS OF STRATIGRAPHIC
SEQUENCES—AN EVALUATION
Although the principle of defining orders by
duration has been almost universally followed,
the actual values used in the definitions scatter
widely. Figure 1 shows the definitions used in
key publications since the introduction of the
concept. In the range of 2nd to 3rd order, the
discrepancies at any one boundary are about Figure 1. Duration of orders of stratigraphic
one-half order; for shorter categories, they are sequences as defined by various authors. In
each category, oldest publication is on top. Figure 2. Durations of sea-level cycles of 3rd
Note large differences, particularly in 4th– and 2nd order of eustatic curve of Haq et al.
*E-mail: [email protected]. 6th orders. (1987). Two orders broadly overlap.

q 2004 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected].
Geology; March 2004; v. 32; no. 3; p. 185–188; doi: 10.1130/G20253.1; 5 figures. 185
Figure 3. Proportions of se-
quences terminated by expo-
sure surfaces (black) and by
flooding surfaces (gray) in well-
documented carbonate rocks.
Each bar represents measured
section with number of ana-
lyzed surfaces on left. Top pan-
el: cycles longer than 1 m.y.
Bottom panel: cycles shorter
than 1 m.y. Sources: Buchbin-
der et al. (2000), D’Argenio et al.
(1997, 1999), Egenhoff et al.
(1999), Föllmi et al. (1994), Hill-
gärtner (1999), Immenhauser et
al. (2001, 2004), Minero (1988),
Saller et al. (1993), Strasser and
Hillgärtner (1998), Van Buchem
et al. (2000, 1996), Wendte et al.
(1992), and Wendte and Muir
(1995).
Figure 4. Migrating shelf breaks on Great
Bahama Bank (Eberli and Ginsburg, 1988)
and offshore New Jersey (Greenlee, 1988)
during Cenozoic phases of progradation. A:
Part of migration path of Bahama shelf
break (as I determined by using Western
Geophysical seismic line described in Eberli
and Ginsburg [1988]); TWT—two-way trav-
considered supratidal features, as is meniscus (Posamentier et al., 1992; Posamentier and eltime. B: Fractal analysis of shelf-break mi-
cement if it occurs in supratidal deposits. Allen, 1999), fully developed sequences on gration paths using box-counting technique
Two important criteria were used for data the scale of 105 yr during the Pliocene– (see text). Straight-line trends for box sizes
221–226 and 227, respectively, indicate
inclusion in Figure 3: (1) detailed and specific Pleistocene (e.g., Carter et al., 1991), and fa- power-law relationship between box size
observations had been made on the nature of cies patterns in plan view (e.g., Rankey, and number of boxes required to cover
each discontinuity surface, and (2) the au- 2002). In an even broader perspective, Van curve. This result, in turn, suggests that
thor(s) accepted the possibility that a flooding Wagoner et al. (2003) interpreted sediment ac- paths of shelf-break migration are fractal
features in these limits. Trends on right are
surface can be a sequence boundary. cumulation radiating from a point source as
somewhat irregular because of finite-size ef-
Figure 3 indicates that standard sequences energy-dissipation patterns that are scale- fects (dashed part); trends break down on
bounded by exposure surfaces and parasequ- invariant over the range of centimeters to left because of limited seismic resolution
ences bounded by flooding surfaces are both hundreds of kilometers. (dotted part).
common among cycles of 103–106 yr duration. Tests for fractal properties of long, contin-
This demonstration is at variance with the as- uous sequence successions seem to be lacking damental sequence controls—sea-level and
sumption of the standard model of sequence in the literature. Figure 4 presents two exam- sedimentation rates. (See Fluegeman and
stratigraphy that 3rd-order sequences are ples from small seismic data sets showing mi- Snow [1989]; Hsui et al. [1993], and Harrison
bounded by exposure surfaces, whereas the gration paths of prograding shelf breaks. The [2002] for sea-level rates and Sadler [1999]
higher-order parasequences are bounded by fine structure of these paths was tested by box- for sedimentation rates.)
flooding surfaces. counting (Turcotte, 1997, p. 135). This tech-
nique measures the ‘‘wiggliness’’ of a curve Conceptual Model of Fractal Sequence
SCALE-INVARIANT MODEL OF and tests for fractal properties by covering the Stratigraphy
SEQUENCES entire curve with progressively smaller boxes The experience with the concept of orders
Scale Invariance in the Sedimentary and plotting box size against the number of in sequence stratigraphy leads to the conclu-
Record boxes needed to cover the curve. A straight- sion that they are subdivisions of convenience.
Sedimentation and erosion, the processes line (power-law) relationship in a bilogarithm- Examination of the architecture of sediment
that are ultimately responsible for the sedi- ic plot suggests a fractal. The slope of the line accumulations indicates that many patterns
ment record, operate in the same fashion over is the fractal dimension, a measure of the in- relevant for the sequence model are invariant
a wide range of scales. It is characteristic of tricacy of the pattern. In Figure 4, power-law in a wide range of scales. Moreover, sea-level
hydrodynamics that flow properties are largely relationships were found, suggesting fractal fluctuations and sedimentation rates, two basic
determined by dimensionless ratios, and few properties. Other examples of prograding mar- forcing factors of sequence architecture, prob-
characteristic scales enter in the analysis. De- gins yielded similar results. However, the di- ably have a fractal nature in the time domain.
positional patterns have been found to be scale agnosis is somewhat tenuous because the On the basis of these observations and in-
invariant over a wide range of time and space. power-law region covers less than two orders sights, I propose the following conceptual
Examples include prograding clinoforms of spatial magnitude. model. The pattern of sequences and systems
(Thorne, 1995), systems-tract architecture on Strong evidence for scale invariance in the tracts is scale invariant and a statistical fractal.
meter-scale features formed in hours or days time domain has been observed on two fun- In the arrangement of systems tracts and the

186 GEOLOGY, March 2004

 operate in similar fashion on much shorter The fractal model proposed here predicts that
time scales. However, the formation of rec- the sequence record, like many other natural
ognizable exposure surfaces depends on soil time series, has the characteristics of noise
processes whose rates are considerably lower. with variable persistence and thus variable
Accumulation of organic matter, one of the predictability (see Turcotte, 1997; Hergarten,
fastest soil processes, takes thousands of years 2002, for examples). The model also predicts
to advance significantly (e.g., Birkeland, 1999, that in the sequence record, the effect of or-
p. 215). The rather easily preserved calcretes dered oscillations, such as the orbital cycles,
form at rates of several millimeters per 103 yr generally is subtle and becomes dominant
(e.g., Robbin and Stipp, 1979). Finally, mod- only in special circumstances (see Harrison,
Figure 5. Scale-invariant model of strati- eling of rock-water interaction suggests that 2002, for examples on sea level).
graphic sequences. S sequences, standard significant decrease in bulk-rock d13C requires
type, and P sequences, parasequence type,
are assumed to both occur in about equal
103–104 yr (Yang, 2001). On the basis of these Purpose and Scope of the Model
proportions in 103–106 yr range. Shaded— indications, I assume that the formation of a The model is meant as a conceptual frame-
highstand systems tract; dotted—transgres- preservable exposure record, even in carbon- work to steer future data analysis and to pro-
sive systems tract; black—lowstand sys- ates, takes at least 103 yr. This number is taken vide a basis for statistical characterization of
tems tract.
as the short time limit of the model because sequences. A logical next step is to determine
it assumes that P and S sequences occur at fractal dimensions of important features of se-
about equal proportions. Among sequences quences and explore the limits of the fractal
nature of sequence boundaries, two types are
shorter than 103 yr, P cycles are likely to dom- domain. This insight, in turn, can be used for
about equally likely in the validity range of
inate because exposure surfaces are likely to interpolating between data sets and extrapo-
the model (Fig. 5).
lack the diagnostic features. lating beyond technical limits of observation,
1. The S sequence, or standard sequence,
The upper time limit for the fractal model e.g., in seismic interpretation.
shows the following succession: highstand
is set by lack of data. As Figure 3 includes no In its present form, the model says nothing
tract → sequence boundary → lowstand tract
cycles longer than 10 m.y., it seems prudent about the origin of the fractal patterns but per-
→ transgressive tract → highstand tract. The
to limit the validity range to sequences shorter mits some speculations. Scale-invariant frac-
sequence boundary is a type 1 or type 2
than 107 yr. Beyond this limit set by the da- tals often appear as the result of other fractals
boundary caused by relative sea-level lower-
tabase, tectonics may set yet another limit. (e.g., Hergarten, 2002). Stratigraphic sequenc-
ing (forced regression).
Few margins of basins or continents exist for es are essentially shaped by the interplay of
2. The P sequence, or parasequence, shows
108 yr without significant episodes of deform- rates of change in accommodation and rates
the following succession: highstand tract →
sequence boundary → transgressive tract → ation or magmatism. Consequently, the record of sediment supply (Jervey, 1988; Schlager,
highstand tract. The sequence boundary is a of cycles having a duration on the scale of 108 1993). As both rates show fractal properties,
type 3 boundary, i.e., a flooding surface that yr is largely derived from flooding and expo- it is not surprising that the resulting sequence
overlies marine deposits without demonstrable sure of continental interiors, whereas data on record inherits this attribute. At a more fun-
evidence of terrestrial exposure or forced re- shorter cycles mainly are from ocean margins. damental level, it may be the complexity of
gression (Schlager, 1999). This difference in setting may explain the depositional systems and their tendency to
The crucial distinction between the two characteristics of first-order cycles reported by evolve toward conditions of self-organized
types is that S sequences are bounded by sur- Vail et al. (1991) and Duval et al. (1998). criticality that give rise to fractal features in
faces of forced regression that require a rela- The spatial range of the model is essentially sequence stratigraphy.
tive fall in sea level. P sequences are bounded a consequence of the time limits. Over the
by type 3 sequence boundaries, i.e., flooding range of 103–106 yr, one typically observes ACKNOWLEDGMENTS
sequence packages that range in thickness Supported by the Vrije Universiteit Industrial As-
surfaces that lack evidence of forced regres- sociates in Sedimentology. I thank Stefan Hergarten
sion. Thus, P sequences and their type 3 from 100 to 103 m and extend horizontally for advice on fractals, Heiko Hillgärtner, Gene Ran-
boundaries may be formed by fluctuations in from 101 to 105 m. The larger horizontal ex- key, Peter Sadler, and Georg Warrlich for important
sediment supply or by relative changes of sea tent is a characteristic of nearly all sedimen- reviews, and Sam Purkis for writing a box-counting
tary bodies and reflects the fact that at the sur- program.
level. The type 3 sequence boundary was rec-
ognized as an important type of unconformity face of Earth, vertical gradients generally
exceed horizontal ones by several orders of REFERENCES CITED
by Vail and Todd (1981, their Fig. 3C) and by Birkeland, P.W., 1999, Soils and geomorphology:
Schlager (1999), who proposed the name. magnitude. New York, Oxford University Press, 430 p.
Conservative estimates of the validity range Buchbinder, B., Benjamini, C., and Lipson-Benitah,
of the model are 103–106 yr in the time do- Fractal Sequence Model vs. Ordered S., 2000, Sequence development of late
Cenomanian–Turonian carbonate ramps, plat-
main, 101–105 m in the horizontal directions Hierarchy of Cycles forms and basins in Israel: Cretaceous Re-
of space, and 100–103 m in the vertical. Jus- In many sequence data sets, the impression search, v. 21, p. 813–843.
tification of these ranges is given in the of a hierarchy of cycles is very strong. The Carter, R.M., Abbott, S.T., Fulthorpe, C.S., and
following. model does not imply that this impression is Haywick, D.W., 1991, Application of global
sea-level and sequence-stratigraphic models in
false. It is characteristic of fractals that the Southern Hemisphere Neogene strata from
DISCUSSION AND CONCLUSIONS same pattern is repeated at finer and finer New Zealand, in MacDonald, D.I.M., ed.,
Validity Range of Conceptual Model scales. Consequently, any snapshot of the Sedimentation, tectonics and eustasy: Sea-
The time range of the model can be justified fractal taken at a certain resolution will show level changes at active margins: International
Association of Sedimentologists Special Pub-
as follows. The statistics on flooding vs. ex- a superposition of coarser and finer patterns. lication 12, p. 41–65.
posure boundaries are based on cycles with The crucial difference to an ordered hierarchy D’Argenio, B., Ferreri, V., Amodio, S., and Pelosi,
durations of 104–106 yr. Physical processes of cycles is their lack of characteristic scales. N., 1997, Hierarchy of high-frequency orbital

GEOLOGY, March 2004 187

 cycles in Cretaceous carbonate platform strata: totti, G., Schoepfer, P., Homewood, P., Vah- phy and sea-level changes from seismic stra-
Sedimentary Geology, v. 113, p. 169–193. renkamp, V., Steuber, T., Masse, J.P., Droste, tigraphy, in Illing, L.V., and Hobson, G.D.,
D’Argenio, B., Ferreri, V., Raspini, A., Amodio, S., H.H.J., van Koppen, J., van der Kooij, B., van eds., Petroleum geology of the continental
and Buonocunto, F.P., 1999, Cyclostratigraphy Bentum, E.C., Verwer, K., Hoogerduijn- shelf of northwest Europe: London, Institute
of a carbonate platform as a tool for high- Strating, E., Swinkels, W., Peters, J., for Petroleum, p. 216–233.
precision correlation: Tectonophysics, v. 315, Immenhauser-Potthast, I., and Al Maskery, Vail, P.R., Mitchum, R.M., Todd, R.G., Widmier,
p. 357–385. S.A.J., 2004, Barremian–lower Aptian Qishn J.M., Thompson, S., Sangree, J.B., Bubb, J.N.,
Duval, B., Cramez, C., and Vail, P.R., 1998, Strati- Formation, Haushi-Huqf area, Oman: A new and Hatlelid, W.G., 1977, Seismic stratigraphy
graphic cycles and major marine source rocks, outcrop analogue for the Kharaib/Shu’aiba and global changes of sea level, in Payton,
in De Graciansky, P.C., et al., eds, Mesozoic reservoirs: GeoArabia, v. 9 (in press). C.E., ed., Seismic stratigraphy—Applications
and Cenozoic sequence stratigraphy of Euro- Jervey, M.T., 1988, Quantitative geological model- to hydrocarbon exploration: American Asso-
pean basins: SEPM (Society for Sedimentary ing of siliciclastic rock sequences and their ciation of Petroleum Geologists Memoir 26,
Geology) Special Publication 60, p. 43–51. seismic expression, in Wilgus, C.K., et al., p. 49–212.
Eberli, G.P., and Ginsburg, R.N., 1988, Aggrading eds., Sea-level changes: An integrated ap- Vail, P.R., Audemard, F., Bowman, S.A., Eisner,
and prograding Cenozoic seaways, northwest proach: Society of Economic Paleontologists P.N., and Perez-Cruz, G., 1991, The strati-
Great Bahama Bank, in Bally, A.W., ed., Atlas and Mineralogists Special Publication 42, graphic signature of tectonics, eustasy, and
of seismic stratigraphy: Volume 2: American p. 47–69. sedimentation, in Einsele, G., et al., eds.,
Association of Petroleum Geologists Studies Minero, C.J., 1988, Sedimentation and diagenesis Cycles and events in stratigraphy: Berlin,
in Geology 27, p. 97–103. along an island-sheltered platform margin, El Springer-Verlag, p. 617–659.
Egenhoff, S.O., Peterhänsel, A., Bechstädt, T., Abra Formation, Cretaceous of Mexico: Pa- Van Buchem, F.S.P., Razin, P., Homewood, P.W.,
Zühlke, R., and Grötsch, J., 1999, Facies ar- leokarst, v. 18, p. 385–405. Philip, J.M., Eberli, G.P., Platel, J.-P., Roger,
chitecture of an isolated carbonate platform: Posamentier, H.W., and Allen, G.P., 1999, Silici- J., Eschard, R., Desaubliaux, G.M.J., Bois-
Tracing the cycles of the Latemar (Middle Tri- clastic sequence stratigraphy—Concepts and seau, T., Leduc, J.-P., Labourdette, R., and
assic, northern Italy): Sedimentology, v. 46, applications: SEPM (Society for Sedimentary Cantaloube, S., 1996, High resolution se-
p. 893–912. Geology) Concepts in Sedimentology and Pa- quence stratigraphy of the Natih Formation
Fluegeman, R.H., and Snow, R.S., 1989, Fractal leontology, v. 7, 209 p. (Cenomanian/Turonian) in north Oman: Dis-
analysis of long-range paleoclimatic data: Ox- Posamentier, H.W., Allen, G.P., and James, D.P., tribution of source rocks and reservoir facies:
ygen isotope record of Pacific core V28-239: 1992, High resolution sequence stratigraphy— GeoArabia, v. 1, p. 65–91.
Pure and Applied Geophysics, v. 131, The East Coulee delta, Alberta: Journal of Van Buchem, F.S.P., Chaix, M., Eberli, G.P., Whal-
p. 307–313. Sedimentary Petrology, v. 62, p. 310–317. en, M.T., Masse, P., and Mountjoy, E., 2000,
Föllmi, K.B., Weissert, H., Bisping, M., and Funk, Rankey, E.C., 2002, Spatial patterns of sediment ac- Outcrop to subsurface correlation of the Upper
H.P., 1994, Phosphogenesis, carbon-isotope cumulation on a Holocene carbonate tidal flat, Devonian (Frasnian) in the Alberta Basin (W
stratigraphy, and carbonate-platform evolution northwest Andros Island, Bahamas: Journal of Canada) based on the comparison of Miette
along the Lower Cretaceous northern Tethyan Sedimentary Research, v. 72, p. 591–601. and Redwater carbonate buildup margins, in
margin: Geological Society of America Bul- Robbin, D.M., and Stipp, J.J., 1979, Depositional Homewood, P.W., and Eberli, G.P., eds., Ge-
letin, v. 106, p. 729–746. rate of laminated soilstone crusts, Florida netic stratigraphy on the exploration and the
Greenlee, S.M., 1988, Tertiary depositional se- Keys: Journal of Sedimentary Petrology, production scales: Centres de Recherches
quences, offshore New Jersey and Alabama, v. 49, p. 175–178. Exploration-Production Elf-Aquitaine Mémo-
in Bally, A.W., ed., Atlas of seismic stratig- Sadler, P.M., 1999, The influence of hiatuses on sed- ire 24, p. 225–267.
raphy: Volume 2: American Association of iment accumulation rates: GeoResearch Fo- Van Wagoner, J.C., Mitchum, R.M., Campion,
Petroleum Geologists Studies in Geology, rum, v. 5, p. 15–40. K.M., and Rahmanian, V.D., 1990, Siliciclas-
v. 27, p. 67–80. Saller, A., Armin, R., Ichram La, O., and Glenn- tic sequence stratigraphy in well logs, cores,
Haq, B.U., Hardenbol, J., and Vail, P.R., 1987, Sullivan, C., 1993, Sequence stratigraphy of and outcrops: Concepts for high-resolution
Chronology of fluctuating sea levels since the aggrading and backstepping carbonate correlation of time and facies: American As-
Triassic (250 million years ago to present): shelves, Oligocene, central Kalimantan, Indo- sociation of Petroleum Geologists Methods in
Science, v. 235, p. 1156–1167. nesia, in Loucks, R.G., and Sarg, J.F., eds., Exploration, v. 7, 55 p.
Hardenbol, J., Thierry, J., Farley, M.B., Jacquin, T., Carbonate sequence stratigraphy: Recent de- Van Wagoner, J.C., Beaubouef, R.T., Hoyal,
De Graciansky, P.C., and Vail, P.R., 1998, Me- velopments and applications: American As- D.C.J.D., Dunn, P.A., Adair, N.L., Abreu, V.,
sozoic and Cenozoic sequence chronostrati- sociation of Petroleum Geologists Memoir 57, Li, D., Wellner, R.W., Awwiller, D.N., and
graphic framework of European basins, in De p. 267–290. Sun, T., 2003, Energy dissipation and the fun-
Graciansky, P.C., et al., eds., Mesozoic and Schlager, W., 1993, Accommodation and supply— damental shape of siliciclastic sedimentary
Cenozoic sequence stratigraphy of European A dual control on stratigraphic sequences: bodies: American Association of Petroleum
basins: SEPM (Society for Sedimentary Ge- Sedimentary Geology, v. 86, p. 111–136. Geologists Official Program, v. 12, p. A175.
ology) Special Publication 60, p. 3–14. Schlager, W., 1999, Type 3 sequence boundaries, in Wendte, J., and Muir, I., 1995, Recognition and sig-
Harrison, C.G.A., 2002, Power spectrum of sea lev- Harris, P.M., ed., Advances in carbonate se- nificance of an intraformational unconformity
el change over 15 decades of frequency: Geo- quence stratigraphy—Application to reser- in Late Devonian Swan Hills reef complexes,
chemistry Geophysics Geosystems, v. 3, arti- voirs, outcrops, and models: SEPM (Society Alberta, in Budd, D.A., et al., eds., Unconfor-
cle 1047, doi: 10.1029/2002GC000300. for Sedimentary Geology) Special Publication mities and porosity in carbonate strata: Amer-
Hergarten, S., 2002, Self-organized criticality in 63, p. 35–45. ican Association of Petroleum Geologists
Earth systems: Berlin, Springer-Verlag, 272 p. Strasser, A., and Hillgärtner, H., 1998, High- Memoir 63, p. 259–278.
Hillgärtner, H., 1999, The evolution of the French frequency sea-level fluctuations recorded on a Wendte, J.C., Stoakes, F.A., and Campbell, C.V.,
Jura platform during the late Berriasian to ear- shallow carbonate platform (Berriasian and 1992, Devonian–Early Mississippian carbon-
ly Valanginian: Controlling factors and timing: lower Valanginian of Mount Saleve, French ates of the Western Canada sedimentary basin:
Fribourg, Multiprint SA, 203 p. Jura): Eclogae Geologicae Helvetiae, v. 91, A sequence stratigraphic framework: SEPM
Hsui, A.T., Rust, K.A., and Klein, G.D., 1993, A p. 375–390. (Society for Sedimentary Geology) Short
fractal analysis of Quaternary, Cenozoic– Thorne, J., 1995, On the scale independent shape of Course Notes 28, 255 p.
Mesozoic, and Late Pennsylvanian sea-level prograding stratigraphic units, in Barton, C.B., Yang, W., 2001, Estimation of duration of subaerial
exposure in shallow-marine limestones—An
changes: Journal of Geophysical Research, and La Pointe, P.R., eds., Fractals in petroleum
isotopic approach: Journal of Sedimentary Re-
v. 98, p. 21,963–21,967. geology and Earth processes: New York, Ple-
search, v. 71, p. 778–789.
Immenhauser, A., Van der Kooij, B., Van Vliet, num Press, p. 97–112.
A., Schlager, W., and Scott, R.W., 2001, Turcotte, D.L., 1997, Fractals and chaos in geology Manuscript received 7 October 2003
An ocean-facing Aptian–Albian carbonate and geophysics: Cambridge, Cambridge Uni- Revised manuscript received 5 November 2003
margin, Oman: Sedimentology, v. 48, versity Press, 396 p. Manuscript accepted 12 November 2003
p. 1187–1207. Vail, P.R., and Todd, R.G., 1981, Northern North
Immenhauser, A., Hillgärtner, H., Sattler, U., Ber- Sea Jurassic unconformities, chronostratigra- Printed in USA

188 GEOLOGY, March 2004