FLOOD07643

• County: Statewide
• Basin: Statewide
• Title: Physical and Economic Feasibility of Nonstructural Flood Plain Management Measures
• Date: 3/1/1978
• Prepared For: US
• Prepared By: US Army Corps of Engineers
• Floodplain - Doc Type: Educational/Technical/Reference Information

J
\
\
\
I,
"
"
\
,
(
"
!
I
(
(
[
l
(
I
,
,
1
<:

i
,
I
i
f
(
~
\
(
i
~
(
i
(
I
I
i
i
i
(
I
r
i
I
(
i
i
;
i
;
i
f
i
I
I
(
,
(

Computational Accuracy
Experience in the computation of expected annual damage has shown that there are a
number of factors associated with the computational technique which can affect the accuracy of
the expected annual value. These include: the number and location of discrete points used to
represent the elevation-frequency and depth-damage relationships; the shape of the
relationships; the numerical integration method used to weight the damage values according to
their probability of occurrence; the degree of asymmetry between elevation-frequency curves
in a family of such curves. Often computer programs are written and used to compute expected
annual damage without the realization that these factors can significantly influence the accuracy
of the computation. In this study most of these factors were recognized at the beginning and
steps were taken to preserve the accuracy of the computations, even so, several problems arose
and several adjustments had to be made.

Most of the analyses were made using fourteen points to represent the elevation-frequency
relationship (Figure A-1 data extrapolated to the annual event) and eighteen points to represent
the depth-damage relationships. The maximum number of points allowed by the computer
program was eighteen. These points were adequate for most computations. Some inconsisten-
cies in computed values did, however, develop. As an example see FigureA-31 and Table A-6,
where, with three feet protection and the 2 year flood event at the first floor, the damage
reduced for a flood hazard factor of 12.0 feet was greater than for a flood hazard factor of 8.0
feet. Since there is virtually no damage below the first floor for a 1SNB structure, with and
without the protection, it was hypothesized that the total damage, Figure A-39 and Table A-8,
should increase as the flood hazard factor increased regardless of the event at the first floor and
with three feet protection damage reduced should decrease since the property above the
protection level is flooded more frequently as the flood hazard factor increases and the percent
being protected is less. It was recognized also that the damage computed was more sensitive
with the 2-year event at the first floor than with some less frequent event.

To investigate this, eighteen, instead of fourteen poin.ts, were selected to define the
elevation-frequency relationship (the additional four points were taken in the more frequent
range of events). This resulted in a lowering of the computed value of total expected annual
damage of from 1.0 to 6.0 percent, and a lowering of damage reduced with three feet protection
of less than 10 percent depending upon the flood hazard factor. This analysis was done for flood
hazard factors from 2.0 to 14.0 and a one story, no basement structure. Next, the data were
plotted and some visual smoothing was done and some of the eighteen points were modified
0.1 to 0.2 feet (elevation for a given frequency). This resulted in changes from 0.0 to 4.0 percent
(from the original total values), again, depending upon the flood hazard factor. The changes in
damage reduced were more than for total damage and consequently most of the
inconsistencies noted above were eliminated. Similar computations were made using the
smoothed data and a one story, with basement depth-damage relationship '(2-year event at the
first floor) and the difference was less than 2 percent of the original total values. Computations
were also made with a less frequent event at the first floor, the 10 year event. Using the
smoothed elevation-frequency data (eighteen points) for one story, with and without basement,
the changes from the original total expected annual damage values using the fourteen points
were negligible, less than 1.0 percent. What this analysis points to is the fact that for the 2-year
event at the first floor, using four additional points (eighteen points altogether), locating the

A-6

-

~ --

~----