Page 110 - Carbonate Sedimentology and Sequence Stratigraphy

CHAPTER 6: FUNDAMENTALS OF SEQUENCE STRATIGRAPHY 101

is open for refinement on the exact proportions of the two Origin of scale-invariant fractal pattern
types depending on expanding statistical data bases. It is
likely that the P:S ratio varies with tectonic setting, long- Scale-invariant fractals often appear as the result of other
term eustatic trends, sediment composition (e.g. siliciclastics fractals (Hergarten, 2002). At a first, surficial level, strati-
or carbonates) and other environmental factors. In the ab- graphic sequences are essentially shaped by the interplay
sence of specific constraints, the model assumes a P:S ratio of of rates of change in accommodation and rates of sediment
1. Similarly, if regional or other specific constraints are lack- supply (Fig. 6.11; Jervey, 1988; Schlager, 1993). Both vari-
ing, the model would assume that the shelf-edge trajectory ables show fractal properties: accommodation is largely
would be a fractal with a fractal dimension, D, in the range governed by relative sea-level change and many sea-level
observed in Fig. 6.19 or other, more extensive data bases. curves of the past have been shown to possess fractal prop-
The fractal model delivers estimates of sedimentation erties (Fluegemann and Snow, 1989; Hsui et al., 1993). The
rates or sea-level fluctuations in analogous fashion to sedi- scale-invariant, fractal nature of sediment supply is indi-
ment anatomy. In the absence of case-specific constraints, cated by the long-term decrease of sedimentation rates with
the model assumes sedimentation rates, S’, to decrease pro- increase in time interval (Sadler, 1981; 1999). At a more
portional to the inverse of the square root of the duration of fundamental level, most geologically important processes
sequences, i.e. of sedimentation and erosion are turbulent and the chaotic
nature of turbulence may be one reason for the dominance

S ∝ t −0.5 of scale-invariant, fractal patterns in stratigraphic sequences
and the sediment record in general. Another possible rea-
This formula is based on the first-order trend found by son is the tendency of sedimentary systems to evolve to a
Sadler (1981, 1999) for the decrease of sedimentation rates self-organized critical state (Bak et al., 1987; Hergarten, 2002;
with increasing observation span. The same approach holds Rankey et al., 2002).
for the prediction of sea level. In the absence of specific
constraints, the model assumes that the power of sea-level
ORIGIN OF SEQUENCES
fluctuations, i.e. the square of the amplitude, decreases with
increasing frequency, f, according to the global trend found The authors of the standard model of sequence stratigra-
by Harrison (2002) phy offered not only a conceptual framework for the stratig-
raphy of unconformity-bounded units but also rather firm
P = 10 −4 · f −2 statements about the origin of stratigraphic sequences. Vail
et al. (1977) argue on sedimentologic grounds that strati-
where f is in cycles per year (cpy), and P is the power in
2
m per cpy. It is of the utmost importance to realize that the graphic sequences are essentially caused by relative sea-
model is open for more detailed quantification of any kind level changes; they argue furthermore that most sequences
as long as it is based on statistically valid data and sound are globally synchronous, thus eustasy must be the main
arguments. cause of sequences. Vail (1987, p.1) lists subsidence, climate
Can the model be applied in the absence of unconformi- and sediment supply as major controls but concludes that
ties? The standard model of sequence stratigraphy already “fundamental control of depositional sequences, is, we be-
states that the sequence-bounding unconformities may pass lieve, short-term eustatic changes of sea level superimposed
laterally into correlative conformities . Cycle-stacking pat- on longer-term tectonic changes.” According to Van Wag-
terns and facies analysis (e.g. Goldhammer et al., 1990; oner et al. (1988, p. 39) “Sequences ... are interpreted to
form in response to the interaction between the rates of eu-
Homewood and Eberli, 2000; Van Buchem et al., 2000; Hill-
stasy, subsidence, and sediment supply.” Overall, the au-
gärtner, 1998) have demonstrated that sequence boundaries,
thors of the standard model clearly assumed that sequences
maximum flooding surfaces etc., may be zones of rapid tran-
are shaped by the interplay of eustasy, local tectonic move-
sition rather than distinct surfaces. The scale-invariant mod-
ments and sediment supply. However, they also postulate
els (Fig. 6.16) are applicable in these situations, too. Ma-
that eustasy dominates the sequence record. The main argu-
rine flooding surfaces, in particular, are prone to change into
ment for this assumption is the success in global correlation
flooding intervals because it is impossible for the ocean to
of sequences and the similarity of the estimates in sea-level
erode the seabed at the same time everywhere.
movements (e.g. Vail et al. 1977, p. 84-91; Haq et al., 1987;
The type-3 sequence boundary in carbonates (p.121) is a
Hardenbol et al., 1998; Billups and Schrag, 2002, reproduced
logical consequence of the co-existence of standard sequence
in Fig. 6.22). For the pre-Neogene, however, the sequence-
architecture and parasequence architecture (Fig. 6.16). This
based sea-level curve contains so many events that their
type of boundary was already recognized by Vail and Todd
spacing is close to the resolution of the best dating tech-
(1981, Fig. 3c) and Schlager (1999b). Drowning unconformi-
niques. An elegant experiment by Miall (1992) illustrates
ties are a special case of type-3 boundaries where the car-
the pitfalls of correlating stratigraphic events under these
bonate factory is shut down.
circumstances (Fig. 6.23). Miall’s (1992) results are partic-
ularly relevant if one considers that the best global correla-
tion is found in deep-water sediments whereas the sequence