small so the exceedance curve is approximated by straight line segments. The
value of the discharge at the midpoint of each segment and its incremental time
in percent is then calculated. The representative value of the sediment discharge
is calculated as the geometric mean of the sediment discharges corresponding to
the water discharges that bound the increment. The daily average discharge is
calculated by multiplying the water discharge by the incremental exceedance
fraction and summing all increments. The daily average sediment discharge is
calculated similarly, by summing all results of multiplying the incremental
sediment load by the incremental exceedance fractions. The average annual
sediment yield is the product of the mean daily value times 365 days.
Points of caution
The sediment discharge rating curve is plotted as water discharge (Q) versus
sediment discharge (QS) on a log-log grid. The typical scatter in such plots
demonstrates that sediment discharge is not a simple function of water discharge.
When the water discharge in cfs is plotted versus the sediment concentration in
ppm, scatter is more apparent than when water discharge is plotted versus
sediment discharge in tons/day. This is due to the spurious correlation between
Q and QS resulting from the dependency of QS on Q. The engineer should
investigate and evaluate any regional and watershed characteristics which might
contribute to scatter. This can be accomplished by testing for homogeneity with
respect to season of the year, systematic changes in land use, type of sediment
load, and type of erosive mechanisms. A multiple correlation approach coupled
with good engineering judgement may be employed to establish the dominant
factors influencing historical concentrations. It is important to predict how these
factors might change in the future and how such changes would impact sediment
concentrations and particle sizes.
Additional factors contributing to scatter include washload concentration and
temperature. The percent of the sediment load that is washload influences the
amount of scatter in the data because the washload depends on its availability
from source areas and not upon hydraulics of flow at the point of interest. Also,
as the concentration of fines increases above 10,000 ppm, the transport rate of
sands and gravels increases significantly. Water temperature may cause a
significant variation in transport capacity of the bed material load. Thus water
temperature variations, when coupled with seasonal changes in land use, may
require that separate warm and cold weather sediment discharge rating curves be
used to achieve acceptable accuracy in the calculated results.
It is usually necessary to extrapolate the sediment discharge rating curve to
water discharges well above the range of measured data. Straight-line
extrapolations typically over-estimate sediment load at high discharges.
Extrapolating the relationship for total concentrations does not guarantee the
proper behavior of individual size classes. Typically, the rating curves for finer
size classes tend to flatten with increasing discharge.
59
Chapter 4
Theroetical Basis for SAM.yld Calculations
Previous Page
