9331

• UCREFRP Catalog Number: 9331
• Author: Allred, T. M. and J. C. Schmidt
• Title: Channel narrowing by vertical accretion along the Green River near Green River, Utah
• USFW Year: 1999
• USFW - Doc Type: Geological Society of American Bulletin
• Copyright Material: YES

lished (Fig. 10). The magnitude of narrowing,
which ceased by about 1938, was approximately
5 m. Channel width at the cableway was stable
between 1939 and 1957, but the channel nar-
rowed another 5 m between 1957 and 1962. Nar-
rowing of an additional 5 m occurred between
1%3 and 1993.
These conclusions are based on analysis of 36
measurements made at discharges between 566
and 623 m3/s. The periods of narrowing and sta-
bility were identified by visual inspection of the
data. MemJ channel widths for the three time pe-
riods were significantly different at the a = 0.15
level (P <0.01) using an analysis of the covariant
structure of the data. Channel-width differences
are not an artifact of bias associated with the
magnitude of discharge at the time of each mea-
surement, because the range of discharge values
for each time interval are statistically equal.

. Changes in Hydraulic Geometry at
the Present Cableway

We analyzed changes in hydraulic geometry at
the present cableway in order to evaluate how the
entire shape of the channel has changed. We used
the 2600 measurements to compute these rela-
tionships for the same time periods we identified
above (Table 3). Hydraulic geometry relation-
ships fully describe the average channel geome-
try during a period of time, because each rela-
tionship describes channel shape and mean
section velocity over the full range of measured
discharges.
Between 1930 and 1938, there was a greater
difference between the width of the channel at
low and high discharge than in the two subsequent
periods (Fig. 11). We infer that base flow must
have been typically confmed within emergent
channel bars. The width of the channel at low dis-
charge between 1939 and 1957 was greater than
between 1930 and 1938, but the rate of change in
width with increasing discharge was lower than in
the previous period. Between 1963 and 1993, the
channel was 5 to 7% narrower than it had been in
the immediately preceding period. There are no
statistically significant differences in the hydraulic
geometry relations for the mean depth or mean
velocity for the three periods.

Changes Elsewhere in the Study Area

The magnitude and style of channel narrowing
in the 26.4-km reach where aerial photographs
were analyzed are consistent with the magnitude
and style of changes measured at the old and
present cableways (Fig. 10). Total active-channel
area through the study reach decreased by ap-
proximately 18% between 1938 and 1993, from

i762

ALLRED AND SCHMIDT

.'~

, ~.

TABLE 2. CHANGES IN TWENTIETH-CENTURY FLOOD FREQUENCY
OFTHE GREEN RIVER NEAR GREEN RIVER, UTAH

Time period Mean annual Recurrence intervalt
flow 1.25 yr 2 yr 5 yr
(m3 S-I) (m3 S-l) (m3 S-I) (m3 S-I)

1895-1929' 217 766 1062 1451
1930-1957 156 544 740 969
1963-1996 160 395 586 849

'No data for 1900-1904 and 1906.
tAli recurrence intervals are calculated using the Log Pearson Type III method.

10 yr
(m3s-1)
1699
1100
1020

4.20 x loti to 3.46 X 106 m2 (Table 4). This corre-
sponds to a reduction in average channel width of
28 m, from 159 m to 131 m, which is slightly
greater than the magnitude of narrowing mea-
sured at the present cableway. Although the tem-
poral resolution of these changes is poor because
of the lengthy intervals between some of the
photo series, the temporal pattern of these
changes is similar to those determined from the
analysis of changes at the present cableway.
Thus, it is reasonable to infer that the timing of
channel change throughout the study reach has
been similar to the timing determined at the pres-
ent cableway.
Within the study reach, channel narrowing
was caused by inset floodplain deposits similar
to the deposit on the right bank at the present ca-
bleway (Fig. 9). These inset deposits are ubiqui-

2500000

2000000

w
(!)
II:
~<-
U Ul
CIl E
Ci e
I- g>
Z Cl
W Q)
::iE-S
Ci
w
CIl

1500000

1000000
166

\

500000

o
o

500

tous throughout the reach; they occur in both
straight and curving sections and on both the in-
side and outside of channel bends.
The magnitude of channel narrowing in the
entire study reach exceeds that measured at the
gage; largely because the study reach originally
included numerous channel islands (Fig. 12).
Many of these islands are now attached to the
banks at most discharges, and the secondary
channels that once surrounded these islands have
become constricted andlor completely filled
with sediment. The surface area of secondary
channels decreased by over 50% between 1938
and 1993 (Table 4). These changes represent an
overall simplification in channel form. Van
Steeter and Pitlick (1998) identified similar re-
ductions in a secondary channel area of the Col-
orado River near Grand Junction, Colorado.

1000

1500

2000

DISCHARGE
(m3/s)

Figure 4. Effective discharge curves for three time periods: 1894-1929 (thin line), 1930-1957
(dotted line), and 1%3-1993 (thick line). Note that the largest reduction in effective discharge was
the result of natural cUmate change, and a smaller reduction occurred foUowing dam closure.

Geological Society of America Bulletin, December 1999