Gravel Bars and Bedforms (Element GB}
The migration and convergence of these bedforms
also lead to midchannel bar fo rmation, the third of
the main processes of bar development (Fig. 6.15).
Leddy et al. (1993) also modeled braid bar development in the laboratory, and proposed three
slightly diferent mechanisms of bar development.
('Choldng avulsion'' occurs when a sediment lobe
migrates through an active channel into an area of
flow expansion, where deceleration and flow diver�
gence occur. Part of the flow is diverted into a different course. Typically, this will involve diversion
across an older bar deposit, and a remnant of the
sediment lobe is left as the nucleus of a new
midchannel bar. "Constriction avulsion'' occurs
when a sediment lobe migrates into an area of
channel harrowing, as at a channel junction. Block�
ing of the channel leads to overbank flooding,
crevassing, and diversion of part of the flow into
a different course, leaving the sediment lobe as a
new bar nucleus. «Apex avulsion" occurs as a result
of the lateral and downstream growth of macroforms, accompanied by corresponding cutbank erosion on the opposite side of the channel. Eventually,
this erosion may lead to breakthrough into lower
areas in the channel complex, and partial flow diver�
sion.
Germanoski and Schumm (1993) modeled the effects of adding and removing sediment load from
experimental braided streams, and documented the
subtle but significant changes in fluvial style that
result. In the case of gravel-bed rivers, addition of
sediment load leads to increased braid-bar development and aggradation, whereas removal of load
leads to amalgamation of braid bars and to channel
incision. Changes in sediment load could be very
local effects resulting from changes in the materials
introduced into the channel by bank erosion, or
longer�term, regional changes brought about by tee�
tonic or climatic causes. Base-level changes will also
modify sediment load, as discussed in Sect. 11.2.2.
The different conditions that lead to the fo rmation of gravel sheets and slip-face bedforms have
been discussed earlier (Sect. 5.2.1.2), drawing on the
work of Hein and Walker (1977). In actual streams,
all these processes may be expected to occur simultaneously, or in succession, as the organization of the
channel and bar complex changes, and conditions of
water and sediment discharge vary. Unfortunately,
there is virtually no information on the internal architecture of the different types of bar deposits
formed by these three processes. Lithofacies types
are as described in Chap. 5, and it is anticipated that,
internally, the bars would consist oflenses and wed143
ges of gravel organized between third� and fo urth�
order bounding surfaces. However, at present, we
can only speculate on the geometry and relative ori�
entation of the surfaces and current structures.
Some comments follow on what is known about
ancient GB deposits. An example is illustrated in Fig.
6.16.
Element GB typically forms multistory sheets tens
to hundreds of meters thick. Flat or irregular erosion
surfaces between mesoforms are common. Steeply
dipping channel margins are rarely seen, partly because they tend to be minor parts of a gravelly fluvial
landscape. Actively migrating channels may undercut older bar gravels, producing cutbanks 1-2 m
high, but when filled with later bar gravels of similar
composition and texture, the cut banks may be difficult to identify (Fig. 6.17).
S.A. Smith (1990) defined three styles of GB accretion in the Triassic Budleigh Salterton Pebble
Beds of southern England. Couplets of Gh and St
separated by fo urth-order surfaces represent deposition in topographically segregated parts of a deep
braided channel (Fig. 6.18a). Units of Gp alternating
laterally with Gh and sandstone lithofacies (Fig.
6.18b) represent the variations in gravel sheet
growth described by Hein and Walker (1977), that
S.A. Smith (1990) postulated took place in a relatively deep channel, as evidenced by the height of
some of the Gp sets (3 m). Superimposed Gt wedges
bounded by fo urth-order surfaces represent scour
fi lls (Fig. 6.18c).
Element GB may be interbedded with minor to
predominant sheets or lenses of element SG: sediment gravity flows. Element SB typically comprises
5-10% of even the coarsest gravel succession, and
represents slack-water deposits, such as abandoned
channel fills (minor CH element, where identifiable),
bar-edge sand wedges, and microdeltas (Rust 1972;
Miall 1977), or the deposits of topographically el·
evated parts of a deep gravel river (Fig. 6.18a), such
as parts of the Donjek River, Yukon (Williams and
Rust 1969; Rust 1972; S.A. Smith 1990). These element associations are discussed fu rther and illustrated in Chap. 8. The development of fourth- and
fifth�order lithosomes in gravelly alluvial fans is discussed in Sect. 10.2.
It is now known that channel confluences are sites
of significant scour in gravel-bed rivers (Ashmore
1993; Bridge 1993b; Ferguson 1993; Siegenthaler and
Huggenberger 1993), and that large scoop-shaped
deposits may occur there. As Ashmore (1993, p. 130)
stated, <(They are one of the few places in gravel
braided stream where flow is sufficiently deep to
The migration and convergence of these bedforms
also lead to midchannel bar fo rmation, the third of
the main processes of bar development (Fig. 6.15).
Leddy et al. (1993) also modeled braid bar development in the laboratory, and proposed three
slightly diferent mechanisms of bar development.
('Choldng avulsion'' occurs when a sediment lobe
migrates through an active channel into an area of
flow expansion, where deceleration and flow diver�
gence occur. Part of the flow is diverted into a different course. Typically, this will involve diversion
across an older bar deposit, and a remnant of the
sediment lobe is left as the nucleus of a new
midchannel bar. "Constriction avulsion'' occurs
when a sediment lobe migrates into an area of
channel harrowing, as at a channel junction. Block�
ing of the channel leads to overbank flooding,
crevassing, and diversion of part of the flow into
a different course, leaving the sediment lobe as a
new bar nucleus. «Apex avulsion" occurs as a result
of the lateral and downstream growth of macroforms, accompanied by corresponding cutbank erosion on the opposite side of the channel. Eventually,
this erosion may lead to breakthrough into lower
areas in the channel complex, and partial flow diver�
sion.
Germanoski and Schumm (1993) modeled the effects of adding and removing sediment load from
experimental braided streams, and documented the
subtle but significant changes in fluvial style that
result. In the case of gravel-bed rivers, addition of
sediment load leads to increased braid-bar development and aggradation, whereas removal of load
leads to amalgamation of braid bars and to channel
incision. Changes in sediment load could be very
local effects resulting from changes in the materials
introduced into the channel by bank erosion, or
longer�term, regional changes brought about by tee�
tonic or climatic causes. Base-level changes will also
modify sediment load, as discussed in Sect. 11.2.2.
The different conditions that lead to the fo rmation of gravel sheets and slip-face bedforms have
been discussed earlier (Sect. 5.2.1.2), drawing on the
work of Hein and Walker (1977). In actual streams,
all these processes may be expected to occur simultaneously, or in succession, as the organization of the
channel and bar complex changes, and conditions of
water and sediment discharge vary. Unfortunately,
there is virtually no information on the internal architecture of the different types of bar deposits
formed by these three processes. Lithofacies types
are as described in Chap. 5, and it is anticipated that,
internally, the bars would consist oflenses and wed143
ges of gravel organized between third� and fo urth�
order bounding surfaces. However, at present, we
can only speculate on the geometry and relative ori�
entation of the surfaces and current structures.
Some comments follow on what is known about
ancient GB deposits. An example is illustrated in Fig.
6.16.
Element GB typically forms multistory sheets tens
to hundreds of meters thick. Flat or irregular erosion
surfaces between mesoforms are common. Steeply
dipping channel margins are rarely seen, partly because they tend to be minor parts of a gravelly fluvial
landscape. Actively migrating channels may undercut older bar gravels, producing cutbanks 1-2 m
high, but when filled with later bar gravels of similar
composition and texture, the cut banks may be difficult to identify (Fig. 6.17).
S.A. Smith (1990) defined three styles of GB accretion in the Triassic Budleigh Salterton Pebble
Beds of southern England. Couplets of Gh and St
separated by fo urth-order surfaces represent deposition in topographically segregated parts of a deep
braided channel (Fig. 6.18a). Units of Gp alternating
laterally with Gh and sandstone lithofacies (Fig.
6.18b) represent the variations in gravel sheet
growth described by Hein and Walker (1977), that
S.A. Smith (1990) postulated took place in a relatively deep channel, as evidenced by the height of
some of the Gp sets (3 m). Superimposed Gt wedges
bounded by fo urth-order surfaces represent scour
fi lls (Fig. 6.18c).
Element GB may be interbedded with minor to
predominant sheets or lenses of element SG: sediment gravity flows. Element SB typically comprises
5-10% of even the coarsest gravel succession, and
represents slack-water deposits, such as abandoned
channel fills (minor CH element, where identifiable),
bar-edge sand wedges, and microdeltas (Rust 1972;
Miall 1977), or the deposits of topographically el·
evated parts of a deep gravel river (Fig. 6.18a), such
as parts of the Donjek River, Yukon (Williams and
Rust 1969; Rust 1972; S.A. Smith 1990). These element associations are discussed fu rther and illustrated in Chap. 8. The development of fourth- and
fifth�order lithosomes in gravelly alluvial fans is discussed in Sect. 10.2.
It is now known that channel confluences are sites
of significant scour in gravel-bed rivers (Ashmore
1993; Bridge 1993b; Ferguson 1993; Siegenthaler and
Huggenberger 1993), and that large scoop-shaped
deposits may occur there. As Ashmore (1993, p. 130)
stated, <(They are one of the few places in gravel
braided stream where flow is sufficiently deep to