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

would be unlikely to be greater than 1 or 2 mnUh for use
with PMP design bursts. A design value of 1 mm/h
seems reasonable where no other data are available. A
value of zero is generally too conservative.

. Humid and sub-humid regions of north eastem and
northern Australia: Somewhat higher CL values than for
south eastern Australia may be appropriate, but values
greater than 3 mm/h would be unusual.

. Tasmania: For western Tasmania, catchments are likely
to be saturated, and zero continuing loss is considered
appropriate for design. Loss rates for south eastern
Australia above should apply to eastern Tasmania.

. Arid and semi-arid regions: The few data available
indicate that a slightly higher value of loss rate may be
appropriate than for more humid regions in the south
east of the continent. It is unlikely that this value would
be greater than 3 mnUh.

. Western Australia: For the forested south west region, a
proportional continuing loss of 85% is recommended.
For the remainder of the State, it is considered unlikely
that CL would be greater than 3 mm/h.

I
I

I

I

I
I

4.3 Hydrograph Models

I
I
I

4.3.1 General

In Australia, hydrograph models have been traditionally
classified into unit hydrograph models and runoff routing
models. Discussion on the general application of these
models to the estimation of frequent to Large floods is
provided in Book V, Sections 2 and 3. The following
discussion is speciflcally related to the application of
models to Large to Extreme floods, i.e. those design
situations in which models are used to estimate floods well
beyond the range for which they can be calibrated or their
performance tested against observed floods.
The quality of the modelled flood hydrographs depends
on three components of the modelling process: (i) the basic
model capabilities and constraints, (ii) the quality of the
catchment representation in the model, and (iii) the
appropriateness of the selected parameter values. General
recommendations for these three components in the
context of Large to Extreme flood estimation are provided
separately in Sections 4.3.2 to 4.3.4, but it should be
recognised that the components are closely linked. The
theoretical advantages of a more flexible model that allows
a more accurate representation of the important catchment
features can only be realised if suitable data or design
information exists to identify appropriate model parameter
values.
Sections 4.3.5 and 4.3.6 provide, respectively, more
specific recommendations on selection of parameter values
for runoff routing and unit hydrograph models.

I
I

I

I

I

I

4.3.2 Model Features and Capabilities

The functionalities of a hydrograph modelling package
for Large to Extreme flood estimation can be divided into
basic and enhanced modelling capabilities. The basic
capabilities indicated in (a) below are regarded as essential
for a modelling package that will allow satisfactory
reproduction of runoff response characteristics over a
range of catchments with different features. Small
catchments and catchments with reasonably uniform
characteristics are less demanding in their basic model
requirements. ThEl.. enhanced model capabilities discussed
ill (b) represent desirable model ellha~ments for lbe
future, and for applications in situations where the
complexity of the catchment or the importance of the
results warrants more detailed modelling. The enhanced

I

I
I

-.-.... -_.....-..-.. -. --';:>-.- -....-...-. .---.

models would also form a sounder basis for extrapolation to
extreme events. The importance of these modelling
capabilities is somewhat dependent upon catchment size,
and judgement is required to determine the extent to which
the following issues need to be considered.

(a) Basic model requirements

(i) Representation of catchment routing elements

Significant variation of routing characteristics over the
catchment, particularly in larger catchments, will require at
least a semi-distributed representation of routing elements
in the catchment (see Book V, Section 3). The model
should have the ability to reflect changes in the routing
response of specific elements resulting from modification of
catchment, channel, or storage components.
While there is some evidence of non-linear routing
response over the range of observed floods in most natural
catchments, it is unclear if this effect persists to the range
of Rare to Extreme floods. In this range the routing
response depends on the amount of storage on the
catchment and may approacr. a linear behaviour. The
recommended procedures in 4.3.5 and 4.3.6 are based on
this assumption. The degree of non-linearitY of catchment
behaviour and its effects are discussed by Pilgrim (1986),
together with the background to the recommended
procedures.

(Ii) Spatial variation of rainfall excess

Where it is necessary to apply design rainfalls non-
uniformly across the catchment (see Section 3.10) the
model should be able to represent spatial variations in
rainfall inputs. A semi-distributed rather than a lumped
model is required in most cases, though if the spatial
patterns used in calibration are similar to those adopted for
design events, then a lumped model may provide
reasonable results.

(iii) Distributed output

Flood estimates are often required at different points of
interest within catchment. The model should thus
adequately represent the progressive routing effects
through the catchment, i.e. it should be internally consistent
to allow matching of observed hydrographs at the
catchment outlet and at required internal points. Il should
be noted that some of the simple hydrograph models in
current use (see iv, below) only provide an adequate
representation of internal flows for locations near the
catchment outlet. For other internal locations it may be
necessary to increase the degree of catchment sub-division
(and re-calibrate the model) to conform with the
recommendations for the minimum number of sub-areas
(Book V, Section 3).

(iv) Modelling packages

Examples of modelling packages that satisfy most of
these basic requirements and are readily available and
widely used in Australia are: RAFTS (Willing and Partners,
t9XX), RORB (La",enson and Mein, 1994), URBS (CarroU,
1998), and WBNM (Boyd et aI., 1994). These packages
also offer options for some of the enhanced model
capabilities described below.

(b) Enhanced model capabilities

(i) Separation of overland and channel routing
elements

While the separation of the catchment into overland
flow (or hillslope) elements and major stream (or channel)
flow elements may not be clear-cut, it is desirable for the