NSW Specific Data Info

NSW Office of Environment and Heritage has developed a guide to assist councils and consultants undertaking studies under the NSW Floodplain Management Program to transition to Australian Rainfall and Runoff 2016.

As part of this transition a study (Review of ARR Design Inputs for NSW report) was undertaken to review and advise on addressing under-estimation bias being experienced when using standard ARR 2016 design event methods with default data from the ARR data hub.

The outcomes of this study indicated that there is significant bias in the standard ARR 2016 design event method with default ARR data hub losses and pre-burst.

It identified that default continuing losses from the ARR data hub over-estimated losses and therefore were not fit for purpose and should only be used where better information was not available (see hierarchy below). If default continuing losses from the ARR datahub are to be used these should only be used with a multiplier of 0.4 applied.

It also found that in cases 4 and 5 in the hierachy below the probability neutral burst initial loss values provided from the ARR data Hub for catchments in NSW should be used in all instances unless a detailed Monte Carlo assessment of pre-burst and losses has been carried out.

Considering this new information, practitioners undertaking flood investigations in New South Wales should use a hierarchical approach to loss and pre-burst estimation. This hierarchy goes from 1 (most preferred) to 5 (least preferred) as indicated in Table 1 and described below.

  1. Use the average of calibration losses from the actual study on the catchment if available.
  2. Use the average calibration losses from other studies in the catchment, if available and appropriate for the study.
  3. Use the average calibration losses from other studies in the similar adjacent catchments, if available and appropriate for the study.
  4. Use the NSW FFA-reconciled losses (See Map & Appendix C Table C3) available through the ARR Data Hub. These losses may be used within the catchment in which they were derived (available through the ARR Data Hub) or similar adjacent catchments with appropriate scrutiny. This is used with the unmodified ARR Data Hub initial losses which requires the application of additional scrutiny to the balance between initial loss and pre-burst to ensure it is reflective of flood history and observations for the catchment being investigated in the lead-up to events. This is particularly important in catchments of 100 km2 or less.
  5. Use default ARR data hub continuing losses with a multiplication factor of 0.4. This is used with the unmodified ARR Data Hub initial losses which requires the application of additional scrutiny to the balance between initial loss and pre-burst to ensure it is reflective of flood history and observations for the catchment being investigated in the lead-up to events. This is particularly important in catchments of 100 km2 or less.

Where good local initial loss data is not available (Cases 4 and 5) the probability neutral burst initial loss values determined in the WMAWater 2019 study and available through the ARR datahub should be used in all instances unless a detailed Monte Carlo assessment of pre-burst and losses has been carried out.

The study resulted in the development of the following information that may inform studies in NSW:

  • To use the FFA-reconciled Losses Map click on the black points representing test gauges to see a plot containing the FFA-reconciled and standard method estimates in comparison to at-site FFA. NSW-reconciled loss values are detailed on these plots and in the Tables C1, C2, and C3.
  • For storm initial losses obtained by methods other than the ARR data hub, burst initial losses should be adjusted using the below equation

ILBurst-chosen=ILStorm-chosen×ILBurst-ARR / ILStorm-ARR

Table 1: Hierarchy of Approaches from most (1) to least (5) preferred

Approach Storm Initial Loss Pre-burst (transformational) IL Burst Continuing Loss
1 Average Calibration Not required or back calculated using ILStorm - ILBurst Calculated from Equation 1 above Average Calibration
2 Average Calibration Not required or back calculated using ILStorm - ILBurst Calculated from Equation 1 above Average Calibration
3 Average Calibration Not required or back calculated using ILStorm - ILBurst Calculated from Equation 1 above Average Calibration
4 NSW FFA reconciled initial loss (see ARR Data Hub) Not required or back calculated using ILStorm - ILBurst Probability Neutral Burst Loss available through ARR Data Hub NSW FFA reconciled continuting losses where available (see ARR Data Hub)
5 ARR Data Hub initial loss Not required or back calculated using ILStorm - ILBurst Probability Neutral Burst Loss available through ARR Data Hub NSW FFA reconciled continuting losses where available (see ARR Data Hub)