PCWA-L 467.pdf - PCWA Middle Fork American River Project ...

174 MEASUREMENT OF STAGE &"lD DISCHARGE
MEASUREMENT PROCEDURES DURING RAPIDLY CHANGING STAGE
The preceding discussion on computing the mean gage height of
discharge measurements demonstrated that under conditions of
rapidly changing stage, measurement procedures must be
streamlined, even at the expense of some accuracy. The reduction in
measurement time makes it possible to obtain a gage-height value
that is representative of the measured discharge. Where streams are
uncontrolled, flood rises are more rapid on small streams than on
large streams, because small streams are subject to flash floods that
may rise and faH with sufficient rapidity to produce peak flows of
almost momentary duration. Consequently tbe discussion that follows
distinguishes between the procedures to be foHowed for measuring
large streams and those for smaH streams, during periods of
rapidly cbanging stage. The procedure to be foHowed for measuring
streams whose flow is controlled by hydroelectric powerplants was
discussed on page 140.
CASE A. LARGE STREAMS
During periods ofrapidly changing stage on large streams, the time
consumed in making a discharge measurement may he reduced by
modifying the standard measurement procedure in the following
manner:
1. Use the 0.6-depth method (p. 134). The O.2-depth method (p. 135)
or the subsurface method (p. 136) may he used if placing the
meter at the 0.6-depth creates vertical angles requiring timeconsuming
corrections, or if the vertical angle increases because
of drift collecting on the sounding line.
2. Reduce the velocity-observation time t about 20-30 s.
3. Reduce the number of sections taken to about 15-18.
By incorporating all three of the above practices a measurement
can be made in 15-20 min. If the subsurface method of observing
velocities is used. some vertical-velocity curves win be needed later to
establish coefficients to convert observed velocity to mean velocity.
Carter and Anderoon (1963) have shown that discharge measurements
having 30 verticals, for which the two-point method ofobservation
was used with a 45-s period of observation, win have a standard
error of 2.2 percent (see p. 181-183). That means that two-thirds of
the measurements made using standard procedures would he in error
by 2.2 percent or less. They have also shown that the standard error
for a 25-s period of observation, using the 0.6-depth method with
depth and velocity observed at 16 verticals, is 4.2 percent. The error
caused by using the shortcut method is generally less than the error
to be expected as a result ofthe shifting flow patterns that commonly