FLOOD01888

• County: Statewide
• Title: River Hydraulics
• Date: 10/15/1993
• Prepared By: US Army Corps of Engineers
• Floodplain - Doc Type: Educational/Technical/Reference Information

EM 1110-2-1416
15 Oct 93

a. Stage data. Stage data are the most accurate type
of hydrologic data. Stage measurement is accurate 10
within the amplitude of wind induced gravity waves and
the consistency of the recording device. Experience has
shown that gravity waves are typically about :I: 0.1 foot
in magnitude; see Figure 5-10. Traditional recording
devices, e.g., strip chart recorders and paper tapes, which
were predominant until the early 80's, tended to lose
their accuracy with time. Each month, when the gage
reader changed the tape, the aulomatic and the manual
gage readings were recorded. Usually the difference was
a couple of tenths of a foot although, occasionally, big
discrepancies were found. The recorded readings were
typically then adjusted by a linear relationship with time
to match the manual reading under the assumption that
the error increased gradually with time. The validity of
this assumption may be questionable. These errors,
which may be hidden, have bearing on how well the
model seems 10 match observed data. Another problem
is that gages sometimes lose their datum. Figure 5-11
shows a comparison of the Des Plaines River stages at

421.2

421.0

~
~
-
~

Q)
g' 420.8
~
(/)
o
'00
o
"0
~ 420.6
Q)
:::;:

III
'0
c:
== 420.4

420.2

Lockport with those at Brandon Road Pool, which is
downstream. For 1974, Brandon Road is higher than
Lockport; hence, the Des Plaines River appears 10 be
flowing backwards. Which gage is correct?

(1) Newer gages have electronic recorders and trans-
mit data via satellite. Still, the gages are subject 10 the
similar losses of accuracy with time. Also, satellite
transmissions are subject to large errors which appear as
spikes in the time series. These spikes are easy to dis-
cern, but if they are input to a simulation they are
disastrous.

(2) Finally, point observations, say the 07:00 read-
ing, are often read from the hourly satellite time series.
Since the data may be oscillating (Figure 5-10) is one
point representative of the overall time series?

b. Flow data. Flow is usually a derived, not a mea-
sured quantity. Periodic flow measurements, using
velocity meters, are initially used 10 define a mting curve

<>p

o

o

o

40 80
Time in hrs (Sept. 1986)

120

Figura 5-10. Oscillation of the 1-hour time series from a satallite for the illinois River at Meredosia

5-14

e

.

..

.

.,

e

.
~

.
.

e