FLOOD02997

• County: Statewide
• Basin: Statewide
• Title: Scour and Fill in Steep, Sand-Bed Ephemeral Streams
• Date: 4/1/1978
• Prepared By: University of Missouri-Columbia
• Floodplain - Doc Type: Floodplain Report/Masterplan

-'

w

.r

SCOUR AND FILL IN STEEP, SAND.BED EPHEMERAL STREAMS

563

TABLE l. CALCUUTED FLOW BEHAVIOR FOR
TWO FLOODS ON QUATAL CREEK
Variable Units January 1974 Oecember 1974
d m 0.229 0.344
W m 32.8 32.8
S none 0,0229 0,0229
f none 0.0241 0.0210
f none 0,0482-0,0217 0.0420-0.0189
Figure 5. Stationary V mI, 2.92-4.35 3,84-5.72
wave and antidune B m 14.63 15.54
(breaking w a vel L m U5 -10.0 I 8.33 -IS ,05
geometry, vertical exagg- H,uII; m 0.736- \.43 1.19-2.14
cranoD 2 x. A.....JHraQ. none 0.360-0.428 0.370-0.428
A.u an 27-61 44-92

rensive chan for the January 1974 eVCfi[. bur was not major.
~...laximum bed reworking from the posrflow bed was 73 em in che
study reach and 66 em at cross seaion AA' in Figure 4b.
Net Change in Bed Elevation. Comparison of survey data taken
in November 1973 and March 1975 shows a nee increase in
mean-bed elevation of 0.033 m in (he main channel of the study
reach, although me part shown in Figure 4b suffered net scour.
This change represents 10% of bankfull flow depth, bu, is small
compared [0 the amount of bed reworking during major floods.
Thus, prediCtion of aCtive-channel stability using Schumm's (1961)
criterion may be valid, but it is not confirmed in these short period
observations.

Discussion

It will be argued herein that the scour shown in Figure 4 suggests
that maximum bed reworking was by antidunes. In January 1974,
maximum scour occurred in three troughs subparallel to the flow.
ln natural streams, antidunes frequently occur in trains with widths
smaller than Stream wid,h (for example, Pierce, 1916, p. 42-43). A
[rain of slowly moving antidunes reworks the stream bed along the
axis of the crain. If antidunes are the largest bed fonns formed dur-
ing a flood.. then greatest bed-form scour and fill occurs where an-
tidunes are the largest and a scour-fill trough results. Figure 4a
suggests that three: separate trains of anridunes operated during the:
January 1974 runoff. Figure 4b suggests tha< the December 1974
tunoff had only one such train and that the main part of the flow
occupied the outside of the bend around the souch bar.
Analysis of Flow Bebavior of a Steep Sand-Bcd Scream. Dunes
are the most common bed form in alluvial streams, but their
amplitudes are no' rigorously fixed by a specified sct of flow
parameters (Vanoni, 1971). Ancidunes are less common because
rhey only occur in steepr sand-bed streams, but their maximum
amplitUdes are analytically related to flow parameters (Kennedy,
1961),
The Steep gradienr of Qua<al Creek (2.3%) suggests thar all flood
flows should be in the upper flow regime, that is, anridunes are me
bed form during a flood. The relacively flar bed lcfc aher floods
shows chat no ripples or dunes formed during rhe waning phase. If
ripples or dunes had been formed., they would have been visible on
[he surface, because they form at the lowest flow rates that allow
sediment transport, and subsequent lower discharges during con-
rinued waning do not have sufficient sediment-transport capability
co destroy chern. Also, two scour-cords were found turned
downstream at the surface after rnese twO flow events, which indi-
cates that an upper flow regime prevailed during the end of the
runoff event. The pattern of maximum scour in Figure 4 is not in-
consistent with bed reworking by antidune trains, which suggests
that chey were the latgest bed forms developed during the runoff

lCflGrTUOlHAl SECTrOH

'ROOSTER TAIL' TRAHSVERSE SECTION

~"
.

e:vents in Quatal Creek. Fortunately, flow velocities and antidune
amplltudes can b:: analytically predicted even though flows were
not measured, since such measurement would have required a re-
cording gaging stacion on Quacal Creek.
I have discussed the relevant equations (Foley, 1977) relating an-
tidune. amplitude A, stationary wave height H, stationary wave and
antidune lortgitudinal wavelength L, rooster tail transverse
wavelength B, mean flow depth d, mean flow velocity V, and chan-
nel slope S (Fig. 5). These equations are summarized below:

v= (8g;s)',

where f = Oarcy-Wcisbach friCtion faCtor, r hydraulic radius
= channel cross-sectional area/wetted perimeter, and g = accel-
eration of gravity.

1 s flf' s 2
0.9sflfs2.

(2a)
(2b)

where f' = equivalent pipe friction faCtor for grain roughness of
the scdimenr in the scream bed (Moody, 1944).

v=(*)'

For antidunes where B >> L.

v=(;;)' [l+(~)'r

H_ = 0.1421.
~ = V, (I _ 2T'd)
H L .

References for the equations are Kennedy (1971; eq. I), Taylor and
Brooks (1961; eq. 2a), Kennedy (1961; eqs. 2a, 3, 4, 5); No,din
(1964; eqs. 2b. 3), and Hand (1969, eq. 6). Using equacions 1, 2b,
4,5, and 6, the range of maximum possible antidune amplitudes
can be determined for a given channel by assuming bankfull flow
and breaking stationary waves (that is, stationary waves of
maximum height).
january 1974 Runoff Event. For given channel geometry and
bed materia~ equations 1 through 6 show thanhe shallow<< of twO
flows has a higher relative roughness and hence a higher friction
facror. This means that the shallower flow has a lower velocity and
smaller anridunes. For bankfuU stage, the shallowest (hence most
conservative) case occurs when there is no mean-bed scour and
when flow dep,h is ,he bankfull dep,h of ,he undisturbed channel
(Fig. 5), Table llisrs the results of calculacions using equations 1
through 6 for the undisturbed channel bed a< cross scerion AA' of
Figure 4a.
In calculacions of Table 1, flow depth was used instead of hy-

!i
;i
"

I

"
,

,
j'
.:;

(I)

':1
'['
, ,

ij:i
.":!i:
'Ii'
,Ii
II
'1'1
I
J
i
t
,
!

(3)

(4)
(5)
(6)

,; I

"
t'
"
, I
I
:
i
i
!

111
: ll\