Avulsion in Fluvial Systems and Its Effect on Alluvial Stratigraphy
323
tions by him regarding alluvial architecture and proposals for the nature of the architectural response to
various combinations of autogenic and allogenic
controls. The eight models that he illustrated in that
paper formed the conceptual basis for all the later
work in this area (Sect. 10.4), including the modern
work on the architecture of fluvial deposits in a
sequence-stratigraphic context (Chap. 13).
A well-known example of an early study of
alluvial architecture is that by Campbell (1976),
who described the predominantly sheet-sandstone
bodies of the Westwater Canyon Member of the
Morrison Formation in northern New Mexico. Some
authors have interpreted this unit as the deposit of a
giant sandy alluvial fan (e.g., Galloway 1980).
Campbell's (1976) reconstruction is illustrated in
Fig. 10.12. The apparent westward migration of the
channels with time caught the attention of many
sedimentologists, and the deposit has been compared to the Kosi fan of India (Miall 1981a; Collinson
1986), which shows a similar westward migration
(Fig. 10.6). However, a recent detailed remapping of
this area by Cowan (1991) demolished this reconstruction and proposed a quite different fluvial style
for the beds (Fig. 10.!3). Cowan (1991) was able to
show that Campbell's (1976) mapping of individual
sandstone sheets was based on outcrop color differentiation. He was able to demonstrate that the different colors of the sandstone bodies were formed by
diagenetic leaching, which was only partly related to
depositional architecture. The curved channel margins that make Campbell's reconstruction so appealing are seen to be mainly inferred. Cowan1s (1991)
new map of the extensive cliff outcrops shows that
the Westwater Canyon Member in fact consists of
laterally extensive sandstone sheets that commonly
have indistinct, fe athered margins and contain considerable internal complexity, resulting from the
migration of bar complexes in a braided fluvial environment. Internal channel scour was particularly
important in this river system. As discussed in Sect.
8.2.14, these deposits are now regarded as an example of a distinctive fluvial style, that of the "highenergy, sand-bed braided river."
On a smaller, intrachannel scale, the stacking of
channels and macroforms in braided systems is largely the product of autogenic avulsion of minor and
major channels triggered by channel and bar aggradation. For example, Fig. 10.14 illustrates stacked
sandstone sheets in the type section of the Castlegate
Sandstone (Upper Cretaceous), near Price, Utah.
Most sheets are composed of laterally and vertically
amalgamated macroforms that were built by lateral
to downstream accretion. As far as can be determined, both the stacking of macro forms and, at the
next level, the stacking of sandstone sheets represent
autogenic migration and switching (avulsion) of
major and minor channels on the braidplain.
Through the limited vertical thickness of this outcrop there are no discernible systematic changes in
channel or bar style, or in orientation, that would
point to some external control on sedimentation.
Autogenic stacking of macroforms was explored
in more detail in another outcrop of this unit, at
Tusher Canyon, near Green River, Utah, where detailed paleocurrent analysis of cross-bedding and
accretion-surface orientations permitted the reconstruction of three of the bar units in a vertical succession. This example is illustrated in Sect. 9.5.7 (Fig.
9.59).
The example of the Castlegate Sandstone illustrates the existence of autogenic processes on two
scales, the intrachannel switching of bars and their
resulting macroform deposits, and the avulsive
switching of the channels themselves, typically as a
result of aggradation. Campbell (1976) used a comparable subdivision of sandstone bodies in his architectural reconstruction of the Westwater Canyon
Member in New Mexico, the larger-scale features of
which have now been shown to be in error, as discussed above. His reference to sandstone sheets as
the deposits of «channel systems" composed of
stacked ·and amalgamated channel deposits has,
however, been confirmed by Godin (1991), although
in modified, reinterpreted fo rm. Wizevich (1993)
performed a similar type of cycle analysis on the Lee
Formation (Pennsylvanian) of Kentucky.
An example of Godin's (1991) detailed vertical
profiles through the Westwater Canyon Member is
illustrated in Fig. 10.15. Small-scale fi ning-upward
cycles 1-6 m in thickness are amalgamated into four
major cycles over the total 41.9 m thickness of the
entire section. This section constitutes a complete
proflle through a mappable middle submember of
the Westwater Canyon Member. Godin (1991) defined three types of autogenic cycle, based on analysis of this and other profiles in the Gallup area of New
Mexico (Fig. 10.16). The small-scale cycles, bounded
above and below by surfaces of third order or higher
rank, he interpreted as the fi ning-upward successions formed by macroform accretion or scour fill.
Channel-scale cycles range from 1 to 10 m in thickness and are composed of stacked macroform cycles.
They are bounded by surfaces of fifth order or higher
rank. These cycles are not distinguished in Fig. 10.15.
Sheet-scale cycles are 4-16 m in thickness (heavy
323
tions by him regarding alluvial architecture and proposals for the nature of the architectural response to
various combinations of autogenic and allogenic
controls. The eight models that he illustrated in that
paper formed the conceptual basis for all the later
work in this area (Sect. 10.4), including the modern
work on the architecture of fluvial deposits in a
sequence-stratigraphic context (Chap. 13).
A well-known example of an early study of
alluvial architecture is that by Campbell (1976),
who described the predominantly sheet-sandstone
bodies of the Westwater Canyon Member of the
Morrison Formation in northern New Mexico. Some
authors have interpreted this unit as the deposit of a
giant sandy alluvial fan (e.g., Galloway 1980).
Campbell's (1976) reconstruction is illustrated in
Fig. 10.12. The apparent westward migration of the
channels with time caught the attention of many
sedimentologists, and the deposit has been compared to the Kosi fan of India (Miall 1981a; Collinson
1986), which shows a similar westward migration
(Fig. 10.6). However, a recent detailed remapping of
this area by Cowan (1991) demolished this reconstruction and proposed a quite different fluvial style
for the beds (Fig. 10.!3). Cowan (1991) was able to
show that Campbell's (1976) mapping of individual
sandstone sheets was based on outcrop color differentiation. He was able to demonstrate that the different colors of the sandstone bodies were formed by
diagenetic leaching, which was only partly related to
depositional architecture. The curved channel margins that make Campbell's reconstruction so appealing are seen to be mainly inferred. Cowan1s (1991)
new map of the extensive cliff outcrops shows that
the Westwater Canyon Member in fact consists of
laterally extensive sandstone sheets that commonly
have indistinct, fe athered margins and contain considerable internal complexity, resulting from the
migration of bar complexes in a braided fluvial environment. Internal channel scour was particularly
important in this river system. As discussed in Sect.
8.2.14, these deposits are now regarded as an example of a distinctive fluvial style, that of the "highenergy, sand-bed braided river."
On a smaller, intrachannel scale, the stacking of
channels and macroforms in braided systems is largely the product of autogenic avulsion of minor and
major channels triggered by channel and bar aggradation. For example, Fig. 10.14 illustrates stacked
sandstone sheets in the type section of the Castlegate
Sandstone (Upper Cretaceous), near Price, Utah.
Most sheets are composed of laterally and vertically
amalgamated macroforms that were built by lateral
to downstream accretion. As far as can be determined, both the stacking of macro forms and, at the
next level, the stacking of sandstone sheets represent
autogenic migration and switching (avulsion) of
major and minor channels on the braidplain.
Through the limited vertical thickness of this outcrop there are no discernible systematic changes in
channel or bar style, or in orientation, that would
point to some external control on sedimentation.
Autogenic stacking of macroforms was explored
in more detail in another outcrop of this unit, at
Tusher Canyon, near Green River, Utah, where detailed paleocurrent analysis of cross-bedding and
accretion-surface orientations permitted the reconstruction of three of the bar units in a vertical succession. This example is illustrated in Sect. 9.5.7 (Fig.
9.59).
The example of the Castlegate Sandstone illustrates the existence of autogenic processes on two
scales, the intrachannel switching of bars and their
resulting macroform deposits, and the avulsive
switching of the channels themselves, typically as a
result of aggradation. Campbell (1976) used a comparable subdivision of sandstone bodies in his architectural reconstruction of the Westwater Canyon
Member in New Mexico, the larger-scale features of
which have now been shown to be in error, as discussed above. His reference to sandstone sheets as
the deposits of «channel systems" composed of
stacked ·and amalgamated channel deposits has,
however, been confirmed by Godin (1991), although
in modified, reinterpreted fo rm. Wizevich (1993)
performed a similar type of cycle analysis on the Lee
Formation (Pennsylvanian) of Kentucky.
An example of Godin's (1991) detailed vertical
profiles through the Westwater Canyon Member is
illustrated in Fig. 10.15. Small-scale fi ning-upward
cycles 1-6 m in thickness are amalgamated into four
major cycles over the total 41.9 m thickness of the
entire section. This section constitutes a complete
proflle through a mappable middle submember of
the Westwater Canyon Member. Godin (1991) defined three types of autogenic cycle, based on analysis of this and other profiles in the Gallup area of New
Mexico (Fig. 10.16). The small-scale cycles, bounded
above and below by surfaces of third order or higher
rank, he interpreted as the fi ning-upward successions formed by macroform accretion or scour fill.
Channel-scale cycles range from 1 to 10 m in thickness and are composed of stacked macroform cycles.
They are bounded by surfaces of fifth order or higher
rank. These cycles are not distinguished in Fig. 10.15.
Sheet-scale cycles are 4-16 m in thickness (heavy