Geology Reference
In-Depth Information
Discharge
Time Series
Water
Suspended
Sediment:
2 mm/yr
monsoon
Fig. 7.5
Discharge time series under a monsoonal rainfall regime.
Water discharge during the summer Indian monsoon dominates the annual flow in the Nepalese Himalaya. Storms
a few days in length create peaks in the monsoonal discharge. High suspended sediment discharge is typically
related to landsliding events that create very peaked sediment fluxes and account for
>
95% of the total sediment
discharge. Integration of the suspended sediment load defines a minimum erosion rate of
∼
2 mm/yr across the
catchment. Modified after Gabet
et al.
(2008).
and can have a strong seasonal dependence.
Input of sediment from adjacent hillsides can
vary in predictable ways in different climatic
regimes. For example, most landslides occur
under saturated conditions. In many areas, satu-
ration requires sustained rainfall prior to the trig-
gering storms. In some locales, such as Taiwan,
the intensity of typhoon precipitation is often
sufficient to trigger landslides without significant
prior storms. In parts of the Himalaya, not only is
there a daily rainfall threshold, but nearly a meter
of rainfall is needed in the prior months before
any significant landsliding occurs (Gabet
et al.
,
2004b). Where landslides drive the large-scale
sediment flux, sediment discharges tend to be
spiky and can be only weakly tied to water dis-
charge (Fig. 7.5). For rivers that are “underloaded”
with respect to their transport capacity, sediment
discharges may peak early in a storm cycle as
material in the bed is mobilized, and then, despite
continuing high water discharge levels, diminish
as all available sediment has been exported.
Similarly, if most sediment input to rivers is
landslide-generated, the stochastic nature of
landsliding can cause a mismatch between water
and sediment discharges (Fuller
et al.
, 2003).
Clearly, because of such variations, it is necessary
to average over a sufficiently long time to be able
to characterize the sediment flux at different river
stages and at different times of the year.
Reliable measurement of the contributions
from the dissolved load, suspended load, and
bedload is only rarely achieved. To estimate
suspended loads accurately, both the vertical and
horizontal sediment concentration profiles, as
well as the spatial distribution of water velocities,
have to be either calculated or measured. Bedload
sampling almost always consists of isolated
measurements at a point in a channel. These
measurements then have to be extrapolated to the
entire stream bed and over the seasonal or flood
cycle. Such extrapolations introduce considerable
uncertainty. Although the flux of bedload material
is generally greatest during flood stage, bedload
sampling during flooding is commonly difficult,
if not impossible. All of these complexities
contribute to the fact that the true bedload flux is
poorly known in most rivers. Bedload is commonly
considered to be
<
10% of the suspended load, but
in steep, heavily loaded rivers, bedload can
exceed 35% (Pratt-Sitaula
et al.
, 2007).
Despite these difficulties, fluvial sediment
fluxes, primarily suspended load only, have been
estimated for many rivers near their entry into
the ocean (Milliman and Meade, 1983; Milliman
and Syvitski, 1992). When the major rivers of the
world are considered, those draining the
Himalaya and the Tibetan Plateau contribute a
disproportionate amount of the sediment
discharge to the world's oceans (Table 7.1).
