FLOOD01798

• County: Statewide
• Title: Australian Rainfall and Runoff 1998, Revision of Book VI - Estimation of Large to Extreme Floods
• Date: 11/28/1998
• Prepared By: Rory Nathan, Sinclair Knight Merz
• Floodplain - Doc Type: Educational/Technical/Reference Information

lJ~rll..l

I
I
I
I
I
I
, I
I
I
I
, I
I,
I
I
I
I
I
I
I

PMP using a transformation based on AEP-neutral
objectives, but its AEP will be smaller.

In the derivation of the PMF, the AEP-neutral objective
for selection of design inputs is explicitly rejected in favour
of adopting conservatively high estimates. With regard to
losses, the general recommendations provided in Section
4.2 should be adopted, i.e. losses should be equal to or
possibly a little less than the minimum value in large floods
observed on the catchment. In ail cases, losses are likely to
be low; in many regions of Australia an initial loss value of
zero and a continuing loss rate of 1 mmlhr will be
appropriate.

Temporal patterns provided with the Generalised PMP
methods could be used (Section 3.9), but note that the
patterns derived for the GSAM method represent smoothed
patterns of average variabimy. Practitioners may care to
adopt more severe patterns, but rearrangement of the
patterns to give the highest possible flood peak is at
variance with the objective of deriving a limiting value of
flood that could reasonably occur.

The hydrograph models used to transform the PMP to
the PMF should follow the genera.1 recommendations
provided in Section 4.3. Parameter values should be
selected in accordance with the recommendations provided
in Sections 4.3.5 and 4.3.6. The selection of other design
inputs, such as initial reservoir level or snowpack depth,
should be representative of the more extreme conditions
that could reasonably be expected to occur.

4.9 ,ncorporat'l"on of Paleohydrological
Estimates

Paleohydrological estimates of floods are based on the
stut of the geomorphic and stratigraphic record of past
flo s, as well as evidence of past floods and streamflow
de Ived from historical, archeological, dendrochronologic,
or ,'other sources. The advantage of paleohydrologic data
(USSR, 1998) is that it is often possible to develop records
that are 10 to 100 times longer than conventional or
historical records from other data sources. This information
thus has the potential to provide estimates of Large to Rare
flood peaks that are independent of rainfall-based
procedures. Such information can provide estimates of
design floods directly, or else can be used to help select
AEP-neutral design inputs for rainfall-based procedures
(Section 4.3.4 c).

Paleoflood data provides estimates on the stage limits
of the largest floods over long time periods. A hydraulic
model is then required to convert the stage information to
estimates of peak discharge. One component of a
paleohydrologic estimate is the time period over which the
peak discharge estimate applies. This information can
provide either exceedance, or non-exceedance bounds on
the flood peaks, which can then be incorporated into a flood
frequency analysis along with maxima derived from other
historic and gauged records. Maximum likelihood
procedures are available that allow the historical record to
be represented by thresholds that were not exceeded, and
by flood events whose magnitude is known only to have
exceeded a threshold, to ~e within a range, or which can be
described by a precise value (Stedinger and Cohn, 1986;
Stedinger and Baker, 1987; Stedinger et aI., 1988).

A detailed description of paleoflood procedures is
outside the scope of this document, though details are
outlined in Pickup (1989), Baker (1987) and Kochel and
Baker (1988). Il should be recognised, however, that the
r$quited analyses require speciaffst interdisciplinary
knowledge (National Research Council, 1988) and
resources that are notionally equivalent to, or greater than

uvvr.. y, ~ .......,,,,,"",,....,'..... I..ClI~C IV LAUCllle: I IV""",...

those required for traditional hydrological investigation. An
increasing number of paleohydrological studies have been
undertaken in Australia to date (e.g. Baker and Pickup,
1987; Baker et aI., 1988; Pickup et aI., 1988; Saynor and
Erskine, 1993; Erskine, 1994; Wahl, 1992; Wahl et aI.,
1994), and the science is increasingly being used as an
input to deriving design flood estimates overseas (e.g.
USBR,1998).

A general review of paleohydrological procedures is
provided in Costa (1986) and Baker et al. (1988), and
Pickup (1989) provides an Australian perspective. One
widely used technique, slackwater studies, uses fine-
grained sediment that accumulates in backwater areas to
construct a detailed history of past floods. Deposition of this
sediment typically requires specific hydraulic conditions that
can be related to maximum flood stage, and usuaily
contains dateable material that can be used to establish the
age of the flood. Other paleoflood techniques make use of
flood scour or trim lines, botanic information such as tree
scars and vegetation patterns, sedimentologic evidence
such as gravel bar heights and gravel sizes, or archeologic
and historical information to establish a maximum flood
stage over some time period (USBR, 1998). .

It is quite possible that rainfall-based and paleohydrol-
ogical procedures will provide different estimates of peak
floods, though this is not surprising given the widely
different sources of information used to derive the
estimates. Efforts should be made to reconcile the
differences. At the very least, design inputs should be
varied within expected limits to see if the differences
between the limits can be narrowed, and ideally the
uncertainties should be explicitly evaluated to determine
the width of the confidence limits. The assumptions behind
each procedure should be carefully examined. For example
with rainfall-based procedures, there is very little known
about the manner in which non-linearity changes with flood
magnitude, and the differences between design fiood
estimates may easily be explained by different assumptions
regarding non-linearity. Similarly, certain assumptions will
be inherent in the adopted paleohydrological technique. For
example assumptions made concerning the stationarity of
climate over the Holocene epoch will have a significant
impact on the manner in which the flood estimates are
treated.

Overall, it is recognised that paleohydrological
techniques have received little attention in Australia to date,
but their potential for providing useful information on Large
to Rare floods has been demonstrated in other countries
(most notably the Western United States). In view of the
potential benefits, it is recommended that the use of
paleoflood data should be considered where expenditure of
the additional resources can be justified.

5 ADDITIONAL DESIGN
CONSIDERA TIONS

5.1 General

There are a number of additional considerations that
are relevant to some design situations and the following
sections detail some of the more common issues that need
to be considered. The importance of these considerations,
and hence the complexity of the techniques required to
adequately address the issues, is very much dependent on
the characteristics of the specific design problem. For
example, Ylhere the storage vatume of a reservoir is large
compared to the volume of catchment runoff, the choice of
initial starting levels in the reservoir is likely to have a more