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

DRAFT D

I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

Credible limit of
extrapolation

~

Large

Rare

Design
rainfall
or flood

Interpolation

Extrapolation

Moderate

Moderate to large

Between
1 in 50 and
1 in 100

Between 1 in 100
and the credible limit
of extrapolation

Book VI - Estimation of Large to Extreme Floods

AEP of PMP varies between
1 in 10' and 1 in 10'

~

~ Event Glass

:1 Nature of proceduresl

Unquantifiable, but nDtionally
very large

:1 Nature of uncertainty I

Beyond the credible
limit of extrapolation

:1 Range of AEP

50

100

2000 10' 105

Annual Exceedance Probability (1 in Y)

10.

Figure 1 Design characteristics of notional design event classes.

the border of the allDwable limits in order to provide a
palatable outcome for the stakeholder. If there are
significant differences in outcome within the range of
uncertainty, then the likely range of consequences must be
explicitly considered in a risk management framework when
develDping mitigation strategies.

The procedures presented herein have been reviewed
by experienced designers and academics from around
Australia. They therefore constitute recommended practice.
Innovation and trialing of new techniques based on
add~ional data and research are strongly encouraged fDr
the estimation Df Large to Rare floods, but an increasing
level of prescription is required as estimates extend beyond
the credible limit of extrapolation. Details concerning the
characteristics of each class of flood are provided below.

(a) Large floods

Large floods are intended to represent those events for
which direct observations relevant to the site of interest are
available. The most common sources of information for this
range of floods are the systematic records of rainfalls or
streamflow, though they include historic information for
notable events that occurred prior to the beginning of
continuous gauged records. Accessible records in general
only extend back to the past 100 years, and thus notionally
the AEPs corresponding to this category are lim~ed to
events more frequent than 1 in 100.

The procedures relevant to the analysis of this type of
information are largely covered by the sections related to
rainfall and flood frequency analysis, and rainfall-runoff
routing (Book 11, Book IV Section 2 and Book V Sections 2
and 3); however, as discussed in Section 1.4, there are a
number of important differences that are specific to
providing estimates that are consistent with Rare to
Extreme floods. The analyses are based on deriving design
flood estimates that lie within the range of direct
observations, and thus are essentially interpolative in
nature. A large body of experience and a great variety of
procedures are available to help the practitioner derive
flood estimates within this range, and the associated

degree of uncertainty in the estimates can be readily
quantified.

(b) Rare floods

Rare floods represent the range of events between the
largest direct observations and the "credible limit of
extrapolation". W~h reference to the latter concept, ~ is
worth noting that the term:

. "credible" is used to represent the limit of extrapolation
that can be estimated without the use of other
confirming information from an essentially independent
source; and,

. "extrapolation" is used to denote estimates that are
made outside the range of observatiDns that are
available at a single site.

The credible limit of extrapolation is dependent upon the
nature of available data that can be obtained at andlor
transposed to the site Df interest Procedures are often
used which are based on the regional pooling of data (both
in space and time), and the quality of the extrapolation
depends on the strength of the assumptions made. It is
important to realise that ir, any given region the credible
limit of rainfall extrapolation may well differ from the limit
applicable to floods.

The notional credible lim~s of extrapolation for a range
of data types in Australia is shown in Table 1 (modified
after USBR, 1998). This table indicates the lower AEPs
corresponding to both typical and the most favourable
situations, though in most cases the credible AEP limits are
likely to be considerably closer to the typical estimates than
the most favourable bounds. At present in Australia using
established techniques it is unlikely that the credible limit of
extrapolation will exceed 1 in 1 0000 AEP.

The analyses required to extrapolate estimates to the
credible limit require substantial resources and a high level
of specialist expertise, and they are thus generally beyond
the level of resources available to a single study.
Practitioners will usually need to rely on processed
information prepared specifically for the region Df interest,
though once such information has been prepared it is likely