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

value for use with the PMP should be equal to or possibly a
little less than the minimum ILb value in large floods
observed on the catchment. For significantly shorter burst
durations, a zero value for 14 is recommended.

In this context of selecting a design loss, some care and
interpretation may be required in assessing the minimum
value in observed floods. Sometimes an apparently
anomalous value occurs that is appreciably lower than all
other derived values. As this may result from the effects of
data errors, it may be desirable to neglect the anomalous
low value in selecting the minimum value. This is in
contrast to the median value recommended for design
floods of higher probabiiities in Book II Section 3.

Recommendations for specific regions are provided
below:

. Humid and sub-humid regions of south eastern
Australia: Temporal patterns of pre-burst rainfall are
available for this region, and thus the procedures
provided in (c) below should be used. If PMP design
bursts are used directly, an ILb value of zero should be
used for shorter duration design bursts.

. Tasmania: For western Tasmania, catchments are likely
to be saturated, and 100% runoff (i.e. 14 = 0) is
appropriate for design. Loss values for south eastern
Australia above should apply to eastem Tasmania.

. Arid and semi-arid regions: The few data available
indicate that no initial loss should be deducted from the
PMP.

. Western Australia: For the forested south west region,
the following values of IL. are recommended:

Winter: ILb = O.
Summer: ILb = 200 mm from the high absorbing
gravels and sands of the lateritic uplands and zero
from the remainder of the catchment.

For the remainder of the State, 14 = o.

(c) Recommended initial loss values for use
with design storms (IL.)

(i) Derivation of design storms from design bursts

The design rainfall and temporal pattern information
presented in Section 3 is relevant to design bursts, not
complete storms. Design information on pre-burst rainfall,
i.e. the amount of rainfall that is likely to fall immediately
prior to the design burst, is required to allow the application
of storm losses with the currently available (burst) design
rainfall data. The Bureau of Meteorology has analysed pre-
burst and pre-storm rainfall for the largest observed storms
in the GSAM region and presented the results in the form of
temporal patterns of pre-burst rainfall (Bureau of
Meteorology, 1998'). Il is important to note that these
temporal patterns were derived from an analysis of areal
rainfalls associated with the largest observed storms on
record. A worked example illustrating the use of pre-burst
temporal pattern is provided in Section 6.3.4.
The pre-burst temporal pallerns are a dimensioniess
representation of the rainfall that is likely to occur between
the beginning of the design storm and the start of the
design burst of specified duration. They are applicabie to all
design rainfalls with AE:Ps rarer than or equal to t in 50,
and this specifically includes Extreme rainfall estimates
(such as the PMP) which are based on maximised
observed storms. The pre-burst temporal patterns provide
the necessary information to derive near-complete design
storms. The onlY mi$Sing information is the posterior
component of the storm, though for most studies exclusion
of the trailing rainfalls should not impact greatly on the
results.

~.......~ .. - ........... .",~...", .... "'~':::I'- .... ....".......,.~ . .............

At the time of writing (1998) pre-burst temporal pattern
information is only available for the GSAM region.
Considerable hydrometeorological resources were required
to derive the patterns, and it is expected that similar
analyses could only be undertaken on a regional basis,
where the costs of deriving the patterns could be offset by
the subsequent benefits accruing from a large number of
studies. Practitioners should make use of similar
information for other regions when it becomes available

(Ii) Applicability of design loss values for different
event magnitudes and durations

Unless specific evidence of significant variation of initial
loss with event magnitude or duration has been found in
the region of interest, the storm initial loss values derived
by the procedures in Section 4.2.2, as representative
(median) values from large events, are applicable to flood
estimation over the whole range, from Large floods to the
PMP design flood, and for all durations.

(d) Continuing loss values (CL) for use with
design bursts and design storms

(i) Large and Rare events

The CL values derived by the procedures in 4.2.2 are
based on the analysis of moderate to large events and are
thus directly applicable to events in that range. For CL
values determined by reconciliation with independently
estimated flood estimates (4.2.2 c), the range of application
depends on the credible limit of extrapolation of floods for
the particular design situation.

(ii) Extreme events

CL values in the range from Rare events to the PMP
design event should vary gradually in the same manner as
for initial loss. Equation 7 can be applied to estimate the
loss rate for an AEP of 1 in Y within this range.

(iii) PMP Design Flood

General guidelines regarding the CL values to be used
with PMP design bursts are given below for various regions
of Australia, based on published data or local experience.
With the general nature of the recommendations, it is not
appropriate to delineate precise boundaries of the regions.
Where possible, greatest reliance should be placed on
values derived from several large observed floods on the
catchment of interest, as discussed previously. Given the
tendency of events of greater rainfall intensity to saturate
greater proportions of the catchment, the largest events are
expected to be associated with the smallest loss rates.
Similarly, long duration events can be expected to be
associated with lower CL values than short duration events.
However, as noted above, any anomalously low values,
thought to result from the effects of data errors in the
volume balance computations, should be neglected.

For short duration events, losses are very small
compared with depths of precipitation, and variations in the
value adopted will have little effect on the magnitude of the
resulting ftood. For longer storms, the rate of loss and the
form of loss adopted can have a considerable effect on
estimated floods, particularly on flood volumes, and greater
care is needed in their selection. An example of the
variation of maximum pond level with loss values is given
by Brown (1982).

Recommendations for specific regions are provided
below:

. Humid and sub-humid regions of south eastern
Australia: For catchments considered similar to the
humid and sub-humid regions of Australia, CL values