Design Intensity-Frequency-Duration

2. Chapter 2 of 6. DESIGN INTENSITY-FREQUENCY-DURATION

2.1 Rainfall Intensity

In order to compute design flows, a rainfall intensity and frequency and duration (IFD) relationship is required.

Book II, Section 1, Vol. 1, of ARR (2001) describes the procedures for determining IFD design rainfall information. Two equivalent procedures are described, one algebraic (in Part A of Clause 1.3), the other graphical (in Part B). In both cases, the procedure has up to eight steps, depending on the requirements. The procedures can be programmed using standard spreadsheet software or commonly used computer languages. There are a number of these programs used in Australian practice, all of which require the nine basic parameters obtained from Volume 2 of Australian Rainfall and Runoff.

Steps 1 to 4 from ARR lead to the determination of design rainfall intensities for the basic durations of 6 minutes, and 1, 12, and 72 hours and for the basic ARI of 2 and 50 years.

Steps 5 to 7 extend these intensity values and plots to standard durations:

• 5, 6, 10, 20, 30 minutes, 1, 2, 3, 6, 12, 24, 48 and 72 hours

and standard ARI's:

• 1, 2, 5, 10, 20, 50 and 100 years.

Step 8 fits a sixth degree polynomial to the intensity-frequency duration values. Maps 1 to 6 in Vol. 2 give intensities for ARI = 2 and 50 years and duration equals 1, 12, and 72 hours.

The values given were derived by fitting log-normal distributions to annual series for the various durations, then adjusting (theoretically) the values for ARI < 10 years to the corresponding partial series values. These were based on the use of the Langbein Equations discussed in Bridge Waterways Chapter 2.

In most locations, the skewness of the log transformed data was > 0 so Log Pearson III (LPIII) distributions fit the data better. To enable the log-normal (approximate) values to be converted to LPIII (more accurate) values, "regional" skew coefficients are given on Maps 7a to 7d in ARR Vol. 2. These were obtained by spatial and temporal smoothing of station skew coefficients for the various rainfall durations.

A frequency curve for a duration of 6 minutes can be obtained from empirical equations relating 6 minute intensities to one hour intensities for the basic ARIs of 2 and 50 years. These equations A(3.1) and A(3.2) in ARR include geographically varying empirical coefficients, given in Maps 8 and 9 of Vol. 2.

In summary, nine parameters need to be read off maps contained in Volume 2 of ARR for application of a generalised procedure for determining design rainfall intensity-frequency -duration (IFD) data relevant to a particular location. These nine parameters are:

​​Parameter	Description​
1	2yr ARI 1hr Duration rainfall
2	2yr ARI 12 hr Duration rainfall
3	2yr ARI 72hr Duration rainfall
4	50yr ARI 1hr Duration rainfall
5	50yr ARI 12hr Duration rainfall
6	50yr ARI 72hr Duration rainfall
7	Coefficient of skewness
8	F2 for determining 6 minute rainfall of 2yr ARI
9	F50 for determining 6 minute rainfall of 50yr ARI

The rainfall intensity information required for the estimation of peak flow for small to medium size catchments by using methods such as Rational Method which are largely based on flood frequency analysis of the same set of flow data. The Intensity Frequency Duration (IFD) tables for Perth, Geraldton, Albany, and Port Hedland can be accessed by using the following links.

​​Intensity Frequency Duration Table
Perth	Geraldton	Albany	Port Hedland

 

2.2 Temporal Patterns

More advanced hydrologic models which predict the full shape of the design flow hydrograph require a temporal pattern for design storms. This temporal patterns information are required for Runoff routing methods described in ARR. The design temporal patterns shall be determined in accordance with the methods described in Book II, Section 2 of ARR. It should be borne in mind that the design temporal patterns were derived from a complex method of examination of observed rainfall data. The patterns have been adjusted to ensure a high degree of compatibility with the derived Rainfall Intensity-Frequency-Duration data. The adjustments were made to ensure that rainfall intensities within a pattern for durations shorter than that for the pattern are no more than 10% greater than those given by the IFD data derived from the procedure discussed in Section 2.1.

2.3 Probable Maximum Precipitation

Determination of the probable maximum precipitation (PMP) involves maximisation and transposition of observed rainfall. For catchment areas up to 1000km2 and storm durations up to 6 hours, a procedure for derivation of PMP is contained in Bulletin 53 published by the Bureau of Meteorology in 1994. The procedure applies modified United States data based on research that the rainfall potential in Australia is similar to that in the United States. The data base of extreme storm events in the United States is much larger, thus providing a statistically sounder basis for maximising and transposing extreme events.

For larger areas and long storm durations, the user is referred to the Bureau of Meteorology for more specialist input from their hydrometeorologists.

2.4 Rainfall of ARI between 100 year and the PMP

Probabilistic rainfall intensities for ARI between the 1 in 100 Annual Exceedence Probability (AEP) rainfall and the PMP can be estimated by the interpolation procedure outlined in ARR (Book VI - Estimation of Large and Extreme Floods).

References

Australian Rainfall And Runoff; A Guide to Flood Estimation, Volume 1, 2001.
Australian Rainfall And Runoff; A Guide to Flood Estimation, Volume 2, 1987.