Geoscience Reference
In-Depth Information
(2009) detect a fluvio-pedogentic unit that they suggest
reflects long-term sediment preservation and, hence, for
sediments above, the local depth of maximum scour that
has occurred since genesis of the unit. From a number of
channels, they develop a regional scour envelope curve
(analogous to a regional flood envelope curve) to show
that maximum scour increases with catchment size up to
10 km
2
, beyond which it maintains an asymptotic value
of c. 0.9 m, implying that this represents the limit of the
scour process, regardless of palaeoflood magnitude.
Cross-section
Plan view
1945
1956
1965
1975
1986
13.4.2
Sediment transport in suspension
Irrespective of the point in a flood when bed scour takes
place, the process provides an easily erodible source of
sediment, especially in sand-bed ephemerals. Added to
this is the material that is brought to the stream channel
by overland flow. Together, these produce sediment con-
centrations that have inspired local farmers to describe the
Colorado River and some of its tributaries as 'too thin to
plough and too thick to drink!' (Beverage and Culbertson,
1964).
From the few measurements of suspended sediment
that have been taken for flash floods (e.g. Nordin, 1963;
Lekach and Schick, 1982; Reid and Frostick, 1987;
Sutherland and Bryan, 1990; Alexandrov
et al
., 2009), it
can be shown that the concentration rises along with wa-
ter discharge, as with perennial streams for which there
is, of course, vastly greater information. However, this
is the only similarity. The constant
a
in the relationship
C
1992
1992
Figure 13.16
Dated stages in the development of the channel
of Nahal Hoga, Israel coastal plain, following the imposition of
soil and water conservation measures in the water catchment.
The cross-sections are vertically exaggerated (after Rozin and
Schick, 1996).
antidunes are spatially discrete and extremely transient.
The channel-wide scour and fill patterns indicated by the
scour chains of Leopold, Emmett and Myrick (1966) may
require a different explanation, perhaps involving plane
beds in the upper flow regime.
Powell
et al
. (2006) explore the possibility that depth
of scour (and subsequent fill) in a sand-bed tributary of
Walnut Gulch varies downstream in a fashion that has
pseudo-regularity (Figure 13.19(a)). For the larger floods
of their record, lozenge-shaped zones of maximum scour
appear to have a longitudinal spacing that is about seven
times the channel width. This is tentatively attributed to ei-
ther secondary currents associated with helical flow cells
or large-scale, turbulence-induced, roller eddies that gen-
erate successive zones of flow acceleration and decelera-
tion. A spatial variation of scour and fill is also noted by
Laronne and Shlomi (2007) in distal, braided, gravel-bed
ephemeral channels of the Dead Sea trough. Here, the dis-
turbance of the bed and subsequent deposition are shown
to scale with flood magnitude. However, in the alluvial
stratigraphy of channels in the Arava, to the south, and for
longer intervals (10
3
aQ
b
, where
C
is sediment concentration (measured
in mg/L) and
Q
is water discharge (measured in m
3
/s), is
anywhere between 6 and 4500 times higher in ephemeral
streams. This reflects the fact that, even at low flows, sed-
iment concentrations are often more than five times as
high as at times of
high
flow in perennial systems. On
the other hand, the exponent
b
is usually less than unity
for ephemeral streams, but always greater than unity for
perennial rivers. This suggests that perennial systems are
more responsive to changes in discharge, but only in rela-
tive terms, since the range of concentrations that might be
expected in an ephemeral stream will be anywhere from
35 to 1700 times higher than that expected in perennial
systems (Frostick, Reid and Layman, 1983).
In a 15-year study of the Nahal Eshtemoa, of the north-
ern Negev, Alexandrov, Laronne and Reid (2007) clearly
demonstrate not only the high concentrations reached
by suspended sediment but also that these are differen-
tiated according to the storm type that generates runoff
(Figure 13.20). Winter frontal storms with low 5-minute
=
