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

I
I
I
I
I
I
i

I
!I
I
I
I
I

,

I
I
I
I
I
I
I

ORAFT 0

Book VI - Estimation of large to Extreme Floods

Table 1 Limit of credible extrapolation for different types of data in Austraffa (modified after USSR, 1998).

AI-site gauged flood data
At-site gauged rainfall data
At-site/regional gauged flood data
At-site gauged and paleoflood data (Section 4.9)
Regional rainfall data (Section 3.3.2)
Regional gauged and paleoflood data
Both aleofloDd and rainfall re ional data sets and extra olation

to be appftcable to a large number of different studies.
There is considerable scope for innovation and trialing of
new techniques for this class of estimation, and adoption
after peer review. It is hoped that continued applied
research will reduce the uncertainty of the estimates and
perhaps extend the limit of extrapDlation.

(c) Extreme floods

Extreme floods, the third class, represent the range of
floods where even a high level of expertise cannot reduce
the level of uncertainty substantially, that is the region
which borders on the "unknowable". Estimates of such
events lie beyond the credible limit of extrapolatiDn, but are
hopefully based on our broadest understanding Df the
physical limits of hydrometeorDlogical processes. It should
be recognised that our understanding of catchment
processes is largely based on observations of floods, and
it is possible that a catchment may change its behaviour
when subjected to Extreme rainfalls.
Any extensions beYDnd the credible limit of
extrapolation should employ a consensus approach that
provides consistent and reasonable values for pragmatic
design. The procedures relating to this range of estimates
should be regarded as inherently prescriptive, as without
empirical evidence or scientific justificatiDn there can be no
rational basis for departing from the consensus approach.
The level of uncertainty of these estimates can only be
reduced by long-term fundamental research. Accordingly it
is important that the procedures related to this class of
flODds be reviewed periodically to ensure that any advances
in our unders~a ding of extreme hydrological and
hydrometeorologi I prDcesses are incorporated into
design practice.

1.4 Relationship with Other Sections of ARR

The main focus of this BDDk is on the estimation of
floods rarer than 1 in 100 AEP; its intention is to
supplement the design information provided in other Books
rather than to replace it In particular, the procedures
described in Sections 3.2, 4.2, 4.3 and 4.4 of this Book rely
heavily on the guidelines and design data provided in Book
II (design rainfalls and losses), Book IV Section 2 (flood
frequency analysis) and Book V Sections 2 and 3 (unit
hydrograph and runoff routing methods).

Despite the different focus of this Book, there are a
number of recommendations that relate specifically to the
estimation of floods with AEPs between 1 in 50 and 1 in
1 DO, and thus have the potential to be in conflict with
guidance given in other Books. The additional
recommendations on floods with AEPs from 1 in 50 to 1 in
100 are required to allow a consistent approach tD be
applied over the full flood range covered in this Book, and
to achieve a smooth transition between estimates of Large

and Rare tD Extreme floods. However, the procedures
described in this Book are not intended for application to
the estimation of flDods more frequent than 1 in 50 AEP.
Indeed, if the lowest flood frequency of interest for a
specific application is 1 in 100 AEP, then the practitioner
need not refer to the provisions of this Book.
Specifically, it is considered that the following
procedures are generally not applicable to events equal to
or more frequent than 1 in 100 AEP unless they are
required in conjunction with estimates of Rare to Extreme
floods:

. Temporal patterns: Section 3.9 refers to temporal
patterns that can be used with long duration storms in
sDutheastern Australia (Nathan, 1992; Bureau of
Meteorology, 1998). The unsmoothed form of these
patterns are only applicable to the estimation of events
rarer than 1 in 50 AEP. In addition, these patterns:

were derived from areal design rainfalls whereas
the pattems in Book II Section 2 were derived from
point rainfalls;

always correspond to within-storm bursts, whereas
the patterns in Book II Section 2 are based on part
of storm, whole storm, or possibly multiple storms;

were derived for the Coastal and Inland Zones of
the GSAM region, which do not coincide with the
zones defined for the patterns presented in Book II
SectiDn 2.

The above factors will tend to introduce inconsistencies
with estimates of more frequent floods, and thus their
use should be restricted to the scope of problems
covered in this Book (as summarised in Section 1.1).

. Pre-burst temporal patterns: SectiDn 3.9.1(b) discusses
the use of pre-burst temporal patterns that can be used
in cDnjunction with the storm loss approach (Section
. 4.2.1b). The pre-burst patterns currently available
(Bureau of Meteorology, 1998') are subject to similar
constraints as described in the preceding dot point, and
thus are not applicable to the estimation of more
frequent floods.

. Design storm losses: While the design storm loss
concept (Section 4.2) is potentially applicable to flood
events of all AEPs, its application is dependent upon
the availability of pre-burst temporal patterns. As
discussed in the preceding dot point, the currently
available pre-burst temporal patterns are only
applicable to southeastern Australia fDr the scope of
problems covered in this Book. Thus at present design
storm losses cannot be applied to the estimation of
mDre frequent events. It should also be recognised that
a considerable body of evidence supporting the
recommended values for design burst losses has been
developed (and validated), and it is possible that the