Growth of Present-Day Concepts, 1978-1988
2.4.3.2 Alluvial Basin Architecture
The next step beyond a reconstruction of the channel
style of a river system is an attempt to reconstruct the
basin architecture. As noted by Friend (1983), an
influential set of diagrams was that published by
Allen (1965a) artd reproduced here as Fig. 2.17.
These diagrams show a variety of styles of alluvial
stratigraphy that reflect the patterns of channel
aggradation and migration of various river types. A
great deal has been learned about fluvial styles since
these diagrams were published.
In order to reconstruct alluvial stratigraphy it is
necessary to carry out regional surface or subsurface
correlation of outcrops or well sections. Correlation
is commonly a problem in fluvial deposits because of
rapid lateral facies changes, alack of regional marker
beds, and numerous internal erosion surfaces (channels and interfluve surfaces). Four types of mediumto long-distance correlations can be carried out under certain conditions, however, by focusing on
floodplain deposits (see Sect. 9.5 for a more extensive discussion of these modern mapping methods).
First, magnetostratigraphy is proving to be extremely valuable for providing local chronostratigraphic datum planes, particularly in the more
laterally extensive and possibly less episodically preserved floodplain sequences. Behrensmeyer and
Tauxe (1982) and Behrensmeyer (1987) provided
excellent examples of architectural interpretation
based on this type of control. They were able to
distinguish and relate to each other interbedded
trunk-river and tributary deposits and to calculate
local sedimentation rates.
Second, the same kind of refined lateral control
may be possible using tuff beds. These are rarely
preserved in channel deposits because of the high
transport energy, but may form widespread marker
horizons on the floodplain. Allen and Williams
(1982) provided an example of a unit less than 30 m
thick that can be subdivided into seven allostratigraphic units (these are "chronosomes,, to use the
term suggested by Schultz 1982), permitting a very
detailed reconstruction of architectural evolution.
Paleosols are the third type of floodplain marker
horizon. These can be distinguished from each other
on the basis of petrological and geochemical signatures, and have the added advantage of yielding a
considerable amount of evidence for local environmental conditions, sedimentation rates, and local
flooding and avulsion frequencies (Bown and Kraus
1981, 1987; Kraus 1987). They may even be detectable in seismicMreflection data. Kraus (1987) sug45
gested that channel migration patterns could be analyzed by studying paleosol maturity. The "pedofacies, of the paleosols matures with time, indicating
the extent of exposure and distance from channel
sediment source (Fig. 2.31). These concepts are discussed further in Chaps. 7 and 10. Fourthly, coal
seams have a similar local utility for local lithostratigraphic correlation.
Very few studies of carefully documented regional alluvial architecture have been published.
Those by Behrensmeyer and Tauxe (1982) and
Behrensmeyer (1987) have been referred to above,
and one set of reconstructions is illustrated in Fig.
2.32. Campbell (1976) and Blakey and Gubitosa
(1984) attempted detailed, large-scale architectural
reconstructions. That by Blakey and Gubitosa (1984)
described paleovalley-fill styles, and examined the
causes of architectural variation, such as regional
changes in subsidence patterns. Campbell (1976)
provided a reconstruction of what has been interpreted as a large-scale, braided, alluvial distributary
system, in which he defined large-scale channel belts
and several scales of channel filling. However,
Cowan (1991) has shown that most of Campbell's
(1976) outcrop units were based on weathering characteristics that reflect diagenetic oxidation and
groundwater leaching, not primary depositional architecture (Sect. 10.3.2).
A useful approach to channel classification was
developed by Friend et al. (1979) and Friend (1983).
Friend (1983) defined channels as "elongate depressions in the alluvial surface, with more or less clearly
defined margins or banks between which the river
flow is restricted for most of the year". In many
rivers, there are channels and other types of"hollov v''
of more than one size (tributarie$, crevasse channels,
chute channels, scour hollows), all of which will
show variations in size and morphology. A careful
analysis of these channels is an essential component
of a facies analysis, and Friend ( 1983) provided three
diagrams to assist in this work (Figs. 2.33, 2.34, 2.35).
Note that this classification is strictly descriptive.
No attempt need be made, until the final interpretation, to assign terms such as "braided" or <'anastomosed'' to the deposits. Sheet flood deposits are,
strictly speaking, unchannelized by definition. They
are characteristic of many distal braid-plain deposits, as discussed in Chap. 8. Fixed channels are
typical of anastoinosed rivers. They typically are described as "ribbon" sandstones, and have width/
depth ratios of less than 15. Some ribbons have
<(wings" of sandstone extending laterally from the
top of the channel margin. These are probably levee
2.4.3.2 Alluvial Basin Architecture
The next step beyond a reconstruction of the channel
style of a river system is an attempt to reconstruct the
basin architecture. As noted by Friend (1983), an
influential set of diagrams was that published by
Allen (1965a) artd reproduced here as Fig. 2.17.
These diagrams show a variety of styles of alluvial
stratigraphy that reflect the patterns of channel
aggradation and migration of various river types. A
great deal has been learned about fluvial styles since
these diagrams were published.
In order to reconstruct alluvial stratigraphy it is
necessary to carry out regional surface or subsurface
correlation of outcrops or well sections. Correlation
is commonly a problem in fluvial deposits because of
rapid lateral facies changes, alack of regional marker
beds, and numerous internal erosion surfaces (channels and interfluve surfaces). Four types of mediumto long-distance correlations can be carried out under certain conditions, however, by focusing on
floodplain deposits (see Sect. 9.5 for a more extensive discussion of these modern mapping methods).
First, magnetostratigraphy is proving to be extremely valuable for providing local chronostratigraphic datum planes, particularly in the more
laterally extensive and possibly less episodically preserved floodplain sequences. Behrensmeyer and
Tauxe (1982) and Behrensmeyer (1987) provided
excellent examples of architectural interpretation
based on this type of control. They were able to
distinguish and relate to each other interbedded
trunk-river and tributary deposits and to calculate
local sedimentation rates.
Second, the same kind of refined lateral control
may be possible using tuff beds. These are rarely
preserved in channel deposits because of the high
transport energy, but may form widespread marker
horizons on the floodplain. Allen and Williams
(1982) provided an example of a unit less than 30 m
thick that can be subdivided into seven allostratigraphic units (these are "chronosomes,, to use the
term suggested by Schultz 1982), permitting a very
detailed reconstruction of architectural evolution.
Paleosols are the third type of floodplain marker
horizon. These can be distinguished from each other
on the basis of petrological and geochemical signatures, and have the added advantage of yielding a
considerable amount of evidence for local environmental conditions, sedimentation rates, and local
flooding and avulsion frequencies (Bown and Kraus
1981, 1987; Kraus 1987). They may even be detectable in seismicMreflection data. Kraus (1987) sug45
gested that channel migration patterns could be analyzed by studying paleosol maturity. The "pedofacies, of the paleosols matures with time, indicating
the extent of exposure and distance from channel
sediment source (Fig. 2.31). These concepts are discussed further in Chaps. 7 and 10. Fourthly, coal
seams have a similar local utility for local lithostratigraphic correlation.
Very few studies of carefully documented regional alluvial architecture have been published.
Those by Behrensmeyer and Tauxe (1982) and
Behrensmeyer (1987) have been referred to above,
and one set of reconstructions is illustrated in Fig.
2.32. Campbell (1976) and Blakey and Gubitosa
(1984) attempted detailed, large-scale architectural
reconstructions. That by Blakey and Gubitosa (1984)
described paleovalley-fill styles, and examined the
causes of architectural variation, such as regional
changes in subsidence patterns. Campbell (1976)
provided a reconstruction of what has been interpreted as a large-scale, braided, alluvial distributary
system, in which he defined large-scale channel belts
and several scales of channel filling. However,
Cowan (1991) has shown that most of Campbell's
(1976) outcrop units were based on weathering characteristics that reflect diagenetic oxidation and
groundwater leaching, not primary depositional architecture (Sect. 10.3.2).
A useful approach to channel classification was
developed by Friend et al. (1979) and Friend (1983).
Friend (1983) defined channels as "elongate depressions in the alluvial surface, with more or less clearly
defined margins or banks between which the river
flow is restricted for most of the year". In many
rivers, there are channels and other types of"hollov v''
of more than one size (tributarie$, crevasse channels,
chute channels, scour hollows), all of which will
show variations in size and morphology. A careful
analysis of these channels is an essential component
of a facies analysis, and Friend ( 1983) provided three
diagrams to assist in this work (Figs. 2.33, 2.34, 2.35).
Note that this classification is strictly descriptive.
No attempt need be made, until the final interpretation, to assign terms such as "braided" or <'anastomosed'' to the deposits. Sheet flood deposits are,
strictly speaking, unchannelized by definition. They
are characteristic of many distal braid-plain deposits, as discussed in Chap. 8. Fixed channels are
typical of anastoinosed rivers. They typically are described as "ribbon" sandstones, and have width/
depth ratios of less than 15. Some ribbons have
<(wings" of sandstone extending laterally from the
top of the channel margin. These are probably levee