434
in perennial systems seems to be taken for granted
and has not been specifically tested in many studies.
In a study of a Cretaceous braidplain system in Utah,
Miall (1993) demonstrated a downstream increase in
sand body thickness over a distance of 140 km and
attributed this to a downstream deepening of fluvial
channels as tributary streams merged into fewer,
larger rivers.
12.6.2 Sand Body Architecture
Given a channel form determined by discharge characteristics and bank materials, the main climatic
control on sand body architecture is the . control
exerted by discharge variability on lithofacies
assemblages and bounding surface characteristics
(Fig. 12.2). It is to be expected that lithofacies heterogeneity will increase with increasing seasonal or ·
longer-term variability in discharge, and that the
frequency of minor internal bounding surfaces (second- and third-order surfaces) will likewise increase.
As noted in Sect. 12.2, jones (1977) classified river
hydrography into five basic patterns, reflecting climatic and topographic controls on discharge variability. These variations are undoubtedly reflected
in lithofacies and architectural differences within
fluvial deposits, but the differences are expected to
be subtle and nor amenable to simple interpretation.
As noted earlier, the sedimentological differences
between high-stage and low-stage sedimentation
may provide no dear information on the time gap
(diurnal, seasonal, flood-related, or longer term)
between these stages. Macroforms and channel-fill
assemblages typically display upward fining, but this
can be generated both by accretionary shallowing at
the depositional site and by waning discharge. The
gradual fi lling of a channel or accretion of a macroform under perennial flow conditions generates a
fining-upward succession, but flood cycles also generate upward fining, because most sedimentation
during floods takes place during the waning stages,
following the erosion that occurs during peak runoff. The two causes of upward fining were not at first
clearly distinguished from each other in sedimentological analyses. When first proposed, the South
Saskatchewan model for sandy braided alluvium was
defined largely on the basis of its cyclicity (Miall
1978c), but the studies of such cyclicity that were
cited include both types. One of these studies
described a portion of the Battery Point Formation
(Devonian) of Quebec (Cant and Walker 1976), deposits which have now been interpreted as the prodWhat Does Fluvial Lithofacies Reveal About Climate?
uct of channel migration and macroform aggradation in low-sinuosity rivers (association 3 of
Lawrence and Williams 1987). Another example was
the Devonian distal braidplain deposit in the Canadian Arctic described by Miall and Gibling (1978).
These are now interpreted as flood cycles fo rmed in
an arid environment (Mia!! 1985). The original
South Saskatchewan model of Miall (1978c) therefore includes at least two distinct fluvial styles (see
Sects. 8.2.13 and 8.2.15).
Luttrell (1993) suggested that "it takes a medium
to large-sized river with sufficient depth and width
before long-term sand flats will develop . ... Only
in the large perennial to intermittent rivers do channel-macroform assemblages dominate the alluvial
architecture. Under sustained discharge conditions
well-developed macro forms show a vertical decrease
in both the scale of bedding and grain size." In her
study of the Kayenta Formation (Jurassic) of the
Colorado Plateau area she interpreted proximal
sediments as the deposits of perennial streams
largely on the basis of the abundance of large
macroforms and channel scours indicating water
depths of up to at least 8 m. In their study of Triassic
sediments in New Mexico, DeLuca and Erikkson
( 1989) also proposed distinguishing deposits formed
under perennial flow conditions from those formed
by ephemeral flow on the basis of the occurrence
of lateral-accretion deposits and a comparison with
the Platte- and South-Saskatchewan-type profiles of
Miall (1978c), which were models based partly on
modern perennial rivers (Sects. 8.2.12, 8.2.13).
Other studies of the Kayenta Formation indicate
that the deposits of the upper flow-regime plane-bed
condition, lithofacies Sh, are important in the proximal regions of the depositional system, in southwestern Colorado (Miall 1988b; Bromley 1991a,b).
As discussed later in this chapter (and in Sect.
8.2.16); the presence of sand bodies several meters
thick composed predominantly of this lithofacies is
normally an indicator of flash-flood sedimen�ation,
a style of discharge that occurs most typically in
ephemeral rivers. The Triassic deposits described
by DeLuca and Erlkkson (1989) also contain an
ephemeral stream facies, and their descriptions of
the deposits seem to confirm aridity throughout the
succession. It is suggested that neither Luttrell
(1993) nor DeLuca and Erlkkson (1989) have made a
convincing case for perennial flow in their studies.
In fact, it is quite possible for ephemeral rivers to
develop deep channels with large macroforms that
accrete only spasmodically, during flash floods (Fig.
12.2B). Shepherd (1987) described examples in the
in perennial systems seems to be taken for granted
and has not been specifically tested in many studies.
In a study of a Cretaceous braidplain system in Utah,
Miall (1993) demonstrated a downstream increase in
sand body thickness over a distance of 140 km and
attributed this to a downstream deepening of fluvial
channels as tributary streams merged into fewer,
larger rivers.
12.6.2 Sand Body Architecture
Given a channel form determined by discharge characteristics and bank materials, the main climatic
control on sand body architecture is the . control
exerted by discharge variability on lithofacies
assemblages and bounding surface characteristics
(Fig. 12.2). It is to be expected that lithofacies heterogeneity will increase with increasing seasonal or ·
longer-term variability in discharge, and that the
frequency of minor internal bounding surfaces (second- and third-order surfaces) will likewise increase.
As noted in Sect. 12.2, jones (1977) classified river
hydrography into five basic patterns, reflecting climatic and topographic controls on discharge variability. These variations are undoubtedly reflected
in lithofacies and architectural differences within
fluvial deposits, but the differences are expected to
be subtle and nor amenable to simple interpretation.
As noted earlier, the sedimentological differences
between high-stage and low-stage sedimentation
may provide no dear information on the time gap
(diurnal, seasonal, flood-related, or longer term)
between these stages. Macroforms and channel-fill
assemblages typically display upward fining, but this
can be generated both by accretionary shallowing at
the depositional site and by waning discharge. The
gradual fi lling of a channel or accretion of a macroform under perennial flow conditions generates a
fining-upward succession, but flood cycles also generate upward fining, because most sedimentation
during floods takes place during the waning stages,
following the erosion that occurs during peak runoff. The two causes of upward fining were not at first
clearly distinguished from each other in sedimentological analyses. When first proposed, the South
Saskatchewan model for sandy braided alluvium was
defined largely on the basis of its cyclicity (Miall
1978c), but the studies of such cyclicity that were
cited include both types. One of these studies
described a portion of the Battery Point Formation
(Devonian) of Quebec (Cant and Walker 1976), deposits which have now been interpreted as the prodWhat Does Fluvial Lithofacies Reveal About Climate?
uct of channel migration and macroform aggradation in low-sinuosity rivers (association 3 of
Lawrence and Williams 1987). Another example was
the Devonian distal braidplain deposit in the Canadian Arctic described by Miall and Gibling (1978).
These are now interpreted as flood cycles fo rmed in
an arid environment (Mia!! 1985). The original
South Saskatchewan model of Miall (1978c) therefore includes at least two distinct fluvial styles (see
Sects. 8.2.13 and 8.2.15).
Luttrell (1993) suggested that "it takes a medium
to large-sized river with sufficient depth and width
before long-term sand flats will develop . ... Only
in the large perennial to intermittent rivers do channel-macroform assemblages dominate the alluvial
architecture. Under sustained discharge conditions
well-developed macro forms show a vertical decrease
in both the scale of bedding and grain size." In her
study of the Kayenta Formation (Jurassic) of the
Colorado Plateau area she interpreted proximal
sediments as the deposits of perennial streams
largely on the basis of the abundance of large
macroforms and channel scours indicating water
depths of up to at least 8 m. In their study of Triassic
sediments in New Mexico, DeLuca and Erikkson
( 1989) also proposed distinguishing deposits formed
under perennial flow conditions from those formed
by ephemeral flow on the basis of the occurrence
of lateral-accretion deposits and a comparison with
the Platte- and South-Saskatchewan-type profiles of
Miall (1978c), which were models based partly on
modern perennial rivers (Sects. 8.2.12, 8.2.13).
Other studies of the Kayenta Formation indicate
that the deposits of the upper flow-regime plane-bed
condition, lithofacies Sh, are important in the proximal regions of the depositional system, in southwestern Colorado (Miall 1988b; Bromley 1991a,b).
As discussed later in this chapter (and in Sect.
8.2.16); the presence of sand bodies several meters
thick composed predominantly of this lithofacies is
normally an indicator of flash-flood sedimen�ation,
a style of discharge that occurs most typically in
ephemeral rivers. The Triassic deposits described
by DeLuca and Erlkkson (1989) also contain an
ephemeral stream facies, and their descriptions of
the deposits seem to confirm aridity throughout the
succession. It is suggested that neither Luttrell
(1993) nor DeLuca and Erlkkson (1989) have made a
convincing case for perennial flow in their studies.
In fact, it is quite possible for ephemeral rivers to
develop deep channels with large macroforms that
accrete only spasmodically, during flash floods (Fig.
12.2B). Shepherd (1987) described examples in the