42
largely intact (Fig. 2.26A), but during intervals of
slow subsidence most of the floodplain deposits and
the upper, finer part of the point-bar succession,
may be planed off, and the result would be multistory coarse units that preserve only the evidence of
high-energy channel-floor sedimentation (Fig.
2.26B). This part of the succession may well be interpreted as the product of a braided stream system,
because of the relatively high proportion of sandy
channel deposits, and the paucity of floodplain fines.
The fluctuation in fl uvial style might be attributed to
climatic or tectonic causes. Allen (1978, p. 145) argued that in a basin undergoing a slowly decreasing
subsidence rate this effect could be the cause of
large-scale coarsening-upward (or at least sandierupward) cycles (cycles similar to those in Fig. 2.26B
overlying those more like 2.24a), whereas such cycles
are commonly attributed to progradation caused by
increased differential movement, such as the coarsening-upward successions of Steel and Aasheim
(1978), Steel et al. (1977), and Heward (!978a). The
stacking and separation of channel bodies and preservation of intervening fines are discussed in Chap.
9, and the interpretation of allogenic cyclicity is discussed further in Sect. 1!.3.6.
2.4.3 Fluvial Architecture
2.4.3. 1 Architectural Scale and the Bounding-Surface Concept
As noted by Allen (1983a, p. 249):
"The idea that sandstone bodies are divisible internally
into •packets' of genetically related strata by an hierarchically ordered set of bedding contacts has been exploited
sedimentologically for many years, although not always in
an explicit manner. For example, McKee and Weir (1953)
distinguished the hierarchy of the stratum, the set of
strata, and the coset of sets of strata, bedding contacts
being used implicitly to separate these entities."
In order to trace the development of these ideas, it
is necessary to reach back to the 1960s. Allen (1966)
showed that flow fields in such environments as
rivers and deltas could be classified into a hierarchical order. His hierarchy was designed as an aid to the
interpretation of variance in paleocurrent data collected over various areal scales, from the individual
bed to large outcrops or outcrop groups. The hierarchy consists of five categories, small-scale ripples,
large-scale ripples, dunes, channels, and the integrated system, meaning the sum of the variances
over the four scales. Miall (1974) added the scale of
the entire river system to this idea, and compiled
Historical Background
some data illustrating the validity of the concept
(Fig. 2.!0).
Brookfield (!977) discussed the concept of an
eolian bedform hierarchy and tabulated the characteristics of four orders of eolian bedform elements:
draas, dunes, aerodynamic ripples, and impact
ripples. These four orders occur simultaneously, superimposed on each other. Brookfield showed that
this superimposition resulted in the formation of
three types of internal bounding surfaces. His firstorder surfaces are major, laterally extensive, flatlying, or convex-up bedding planes between draas.
Second-order surfaces are low to moderately dipping surfaces bounding sets of cross-strata fo rmed
by the passage of dunes across draas. Third-order
surfaces are reactivation surfaces bounding bundles
of laminae within cross-bed sets and are caused by
localized changes in wind direction or velocity.
(mesoforms to microforms).
Similar hierarchies of internal bounding surfaces
have been recognized in some subaqueous bedfo rms. A particularly useful approach was that of
jackson (1975). Building on Russian work, he defined three classes of bedforms: microforms (e.g.,
ripples), mesoforms (e.g., dunes, sand waves, produced by ('dynamic events", such as floods), and
macroforms (longer-term geomorphic products,
such as point bars, sand flats, eolian draas). Reactivation surfaces (Collinson 1970) may develop in
bedforms as a result of changes in stage or flow
direction or under conditions of constant stage,
where bedforms randomly interact (McCabe and
jones 1977). They may also separate individual
smaller-scale elements within a larger unit, such as
the component packages of mesoforms within a
macroform. They have been documented in smallscale (Collinson 1970) and large-scale fluvial crossbed sets (Jones and McCabe 1980; Fig. 2.27) and in
tidal sand waves (Allen 1980). In the latter case, the
sand waves rest on horizontal surfaces analogous to
Brookfield's first-order surfaces and contain second-order reactivation surfaces that relate to individual tidal cycles. Third-order surfaces bound
individual cross-bed sets.
Brookfield's (1977) development of the relationship between the time duration of a depositional
event, the physical scale of the depositional product,
and the geometry of the resulting lithesome was a
major step forward that has been of considerable use
in the analysis of eolian deposits. Brookfield (1977),
Gradzinski et al. (1979), and Kocurek (1981) showed
how these ideas could be applied to the interpretation of ancient eolian deposits. Kocurek (1988) has
largely intact (Fig. 2.26A), but during intervals of
slow subsidence most of the floodplain deposits and
the upper, finer part of the point-bar succession,
may be planed off, and the result would be multistory coarse units that preserve only the evidence of
high-energy channel-floor sedimentation (Fig.
2.26B). This part of the succession may well be interpreted as the product of a braided stream system,
because of the relatively high proportion of sandy
channel deposits, and the paucity of floodplain fines.
The fluctuation in fl uvial style might be attributed to
climatic or tectonic causes. Allen (1978, p. 145) argued that in a basin undergoing a slowly decreasing
subsidence rate this effect could be the cause of
large-scale coarsening-upward (or at least sandierupward) cycles (cycles similar to those in Fig. 2.26B
overlying those more like 2.24a), whereas such cycles
are commonly attributed to progradation caused by
increased differential movement, such as the coarsening-upward successions of Steel and Aasheim
(1978), Steel et al. (1977), and Heward (!978a). The
stacking and separation of channel bodies and preservation of intervening fines are discussed in Chap.
9, and the interpretation of allogenic cyclicity is discussed further in Sect. 1!.3.6.
2.4.3 Fluvial Architecture
2.4.3. 1 Architectural Scale and the Bounding-Surface Concept
As noted by Allen (1983a, p. 249):
"The idea that sandstone bodies are divisible internally
into •packets' of genetically related strata by an hierarchically ordered set of bedding contacts has been exploited
sedimentologically for many years, although not always in
an explicit manner. For example, McKee and Weir (1953)
distinguished the hierarchy of the stratum, the set of
strata, and the coset of sets of strata, bedding contacts
being used implicitly to separate these entities."
In order to trace the development of these ideas, it
is necessary to reach back to the 1960s. Allen (1966)
showed that flow fields in such environments as
rivers and deltas could be classified into a hierarchical order. His hierarchy was designed as an aid to the
interpretation of variance in paleocurrent data collected over various areal scales, from the individual
bed to large outcrops or outcrop groups. The hierarchy consists of five categories, small-scale ripples,
large-scale ripples, dunes, channels, and the integrated system, meaning the sum of the variances
over the four scales. Miall (1974) added the scale of
the entire river system to this idea, and compiled
Historical Background
some data illustrating the validity of the concept
(Fig. 2.!0).
Brookfield (!977) discussed the concept of an
eolian bedform hierarchy and tabulated the characteristics of four orders of eolian bedform elements:
draas, dunes, aerodynamic ripples, and impact
ripples. These four orders occur simultaneously, superimposed on each other. Brookfield showed that
this superimposition resulted in the formation of
three types of internal bounding surfaces. His firstorder surfaces are major, laterally extensive, flatlying, or convex-up bedding planes between draas.
Second-order surfaces are low to moderately dipping surfaces bounding sets of cross-strata fo rmed
by the passage of dunes across draas. Third-order
surfaces are reactivation surfaces bounding bundles
of laminae within cross-bed sets and are caused by
localized changes in wind direction or velocity.
(mesoforms to microforms).
Similar hierarchies of internal bounding surfaces
have been recognized in some subaqueous bedfo rms. A particularly useful approach was that of
jackson (1975). Building on Russian work, he defined three classes of bedforms: microforms (e.g.,
ripples), mesoforms (e.g., dunes, sand waves, produced by ('dynamic events", such as floods), and
macroforms (longer-term geomorphic products,
such as point bars, sand flats, eolian draas). Reactivation surfaces (Collinson 1970) may develop in
bedforms as a result of changes in stage or flow
direction or under conditions of constant stage,
where bedforms randomly interact (McCabe and
jones 1977). They may also separate individual
smaller-scale elements within a larger unit, such as
the component packages of mesoforms within a
macroform. They have been documented in smallscale (Collinson 1970) and large-scale fluvial crossbed sets (Jones and McCabe 1980; Fig. 2.27) and in
tidal sand waves (Allen 1980). In the latter case, the
sand waves rest on horizontal surfaces analogous to
Brookfield's first-order surfaces and contain second-order reactivation surfaces that relate to individual tidal cycles. Third-order surfaces bound
individual cross-bed sets.
Brookfield's (1977) development of the relationship between the time duration of a depositional
event, the physical scale of the depositional product,
and the geometry of the resulting lithesome was a
major step forward that has been of considerable use
in the analysis of eolian deposits. Brookfield (1977),
Gradzinski et al. (1979), and Kocurek (1981) showed
how these ideas could be applied to the interpretation of ancient eolian deposits. Kocurek (1988) has