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

downstream from the study reach, preferably at a reliable
buundary condition. If starting conditions are DOt known,
the engineer must ensure that profile computations begin
sufficiently far downstream that any errors in estimating
starling water surface elevation will be eliminated by
profile convergence to the correct elevation downstream
of the study reach. This distance is mainly a function of
the stream slope. Additional guidance on selecting the
correct distance downstream of the study reach is given
in "Accuracy of Computed Water Surface Proftles"
(U.S. Army Corps of Engineers 1986).

(3) The channel n value can be calibrated for various
flows if stage-discharge data are available (e.g. at a
gage). Once a match of computed and actual stages at a
gage site for in-bank flows is obtained, the channel n
may be held constant and the overbank n calibrated for
different historic floods. For one or more known dis-
charges, the computed proftle should be plotted and
compared with measured stages and highwater marl<s. It
should not be expected that the two will exactly coincide.
A successful calibration occurs when the computed pr0-
file is close to the majority of high water marl<s, with
some scatter allowed. Means to achieve a calibration
include changes 10 Manning's n, adjustments 10
expansion/contraction coeffICients where warranted, mod-
ifications to effective flow boundaries, or 10 bridge
geometry descriptions. Typically, most of the adjust-
ments are 10 Manning's n.

(4) Considerable uncertainty exists in the estimation
of n, with estimates by experienced hydraulic engineers
commonly differing by :!:20 percent at the same stream
section (U.S. Army Corps of Engineers 1986). Thus, one
can reasonably justify an increase or decrease of this
magnitude to calibrate a model. The hydraulic engineer
should be cautious if an "unreasonable" adjustment 10 n
is required for calibration. Rigorous guidance on accept-
able calibration errors cannot be given. The judgment
and experience of the responsible hydraulic engineer and
reviewers is foremost. Rules of thumb of:l: I foot are
often used, but this criterion may not be acceplable for
all situations, particularly for steep streams. Some gen-
eral considerations for the calibration process are given
in Table 3-5. Figure 3-3 shows an example of satis-
factory water surface elevation calibration for a stream
reach. The process and rationale for calibration should
be documented in the study reports.

(5) Additional calibration data are necessary for the
application of two-dimensional, unsteady flow, and sedi-
ment transport models. Each chapter on the application
of the various methods provides information on model
calibration and verification.

d. Verification. The last step in the calibration pro-
cess is verificlition of the model. This operation is most
desirable, but is not always possible, often requiring
more data than is available. The verification process is

Table 3-5
Data Gathering/Calibration Coneiderations

Obtain as many highwater marks (HWM) as possibla aIlar any significant flooding. no mattar how close togathar and how
inconsiSlant with nearby HWM's. Physically describe aach HWM location so that surveys may ba obtainad at a latar data.

Obtain highwatar marks upstream and downstream 01 bridges it possibla, so that tha affects causad by thase obstructions
can ba astimatad and so that bridge modeHng procedures may be confinned.

Chock on bridgalculvart debris blockagas with local residents. For urban streams. chack with residents and newspaper filas
on occurrences of bridge opening blockagas by automobilas or dabris.

For historical flooding, chock on land usa changas, both besin wide and local. sinca tha flood(s) occurred.

What has been happening to the stream sinca the last flood? Erosion or deposition that may have occurred sinca historic
floods. if significan~ wiH render calibration with today's channel configuration invalid.

If HWM's are taken from dabris lines. remamber that wave wash can resufi in the debris Hna baing higher than tha HWM.
partiCUlarly for pools.

Is the observer giving tha HWM biased? A homeowner may give an exaggeratad HWM if the owner thinks it might benafit
a project; the owner with a house for sale may give a low estimata or indicata no flooding occurs H halshe thinks it will
affect the sale.

3-12

e

.
"

.
,

e

f

,
.

e