This page provides references to some interesting papers related to ARR2016.
Accuracy of ARR methods
Babister, M., Retallick, M., Podger, S., McLuckie, D. and Frazer, A. (2019) How reliable is your design flood estimate. Presentation to the FMA conference, Canberra, May. http://www.floods.org.au/site/canberra. (link to presentation).
- No validation had been done on the full set of ARR2016 inputs to test for bias
- Analysis of 165 catchments in NSW
- The data hub values of continuing loss (CL) are too high in NSW
- Data hub CL values should be multiplied by 0.4
- Using the median pre-burst gives results in design flood values being underestimated
- New data hub values are available for NSW
Kus, B., O’Loughlin, G. and Stack, B. (2018) Applying ARR2016 to Stormwater Drainage Design. Stormwater 2018. (link).
Abstract: The methods and advice presented in Australian Rainfall and Runoff 2016 are not yet complete, and there has been a slow uptake of these by the stormwater industry. The paper notes the parts of ARR 2016 relevant to urban drainage system design, and assesses the most important of these – the handling of ARR 2016 rainfall ensembles and the application of a new urban hydrological model using effective impervious areas (EIAs) and initial and continuing losses (IL-CL). Using software, analysis and design of drainage systems with ensembles of rainfall patterns is more complicated than with previous rainfall inputs, but is not particularly difficult to apply or interpret. The estimation of EIAs is illustrated by an analysis of storm data from Jamison Park, Penrith. DRAINS models of sub-catchments at various locations show that that the IL-CL model performs similarly to established models such as ILSAX and the extended rational model, but that there are uncertainties in its application. Other matters discussed include (a) the effects of changed design rainfall intensities and climate change adjustments (b) the analysis of stormwater detention storages, and (c) the relevance of the rational method. The parts of ARR 2016 that can be readily adopted by designers are noted, and the need for further data collection and development of methods is emphasised.
Stack, B. and Kus, B. (2018) Mean vs Median: analysing stormwater drainage results with ARR2016 (link).
Abstract: In summary, the median is a more robust indicator of central tendency and this is why we prefer it for urban drainage. With an even number of storms (10) the median peak value is defined as the mean of the peak values for the Rank 5 and Rank 6 storms. DRAINS uses the storm above the median (i.e. Rank 6) so that in fact it is slightly more conservative than simply using the median.
Swan, R., Guest, R., Sommerville, H. and Haywood, J. (2018) ARR2016 – Adopting a practical methodology for catchment scale urban flood mapping projects. Floodplain Management Australia Conference, Surfers Paradise, May 2018 (link to abstract book at FMA website).
Abstract: The recent release of Australian Rainfall and Runoff 2016 presents high-level guidance on several new approaches that have vastly changed the way in which flood modelling is to be undertaken. This has been introduced with very little guidance on how these approaches will be applied practically to catchment scale urban flood modelling projects. These projects are typified by significant variance in the spatial scale of elements that require flood modelling, from the smallest council pipe through to large drains and creeks. Together with Melbourne Water, the new guidance presented in ARR 2016 was incorporated into a modelling methodology which is able to be applied to projects of all scales. The developed methodology takes into account of such key ARR 2016 aspects as defining and differentiating between effective and indirectly connected impervious areas, calculation of initial and continuing losses and selection of appropriate temporal patterns using Monte Carlo analyses. The selection of design events to be taken through to hydraulic modelling becomes a key constraint of these projects. This methodology was applied to two spatially varying catchment scale flood mapping studies in Melbourne with the results compared to those developed using ARR 1987 for the twenty percent and one percent AEP design events. The paper details the issues encountered when adopting the proposed approach and provides practical guidance for engineers and authorities when developing modelling guidelines.
Wood, D. (2018) A review of modelling practice in a gauged urban catchment under ARR2016. Floodplain Management Australia Conference, Surfers Paradise, May 2018 (link to abstract book at FMA website).
Abstract: The recent ARR2016 methodology has resulted in a significant change in the input required from the engineer undertaking the hydraulic assessments. From the selection of the temporal patterns and loss rates to the selection of appropriate hydraulic model parameters and model setup significantly more information is present for the engineer to digest and interpret. In order to assess the range of variation that may occur and discuss the implications, a three-way blind test of hydraulic modelling within a gauged urban catchment was undertaken. Participants were only given the raw underlying data (topography and 1d networks) and tasked to develop a hydraulic model with the aim of comparing the results to a Flood Frequency Analysis (FFA) for the site. No guidance was provided to the participants with regards to modelling method or validation approach and the FFA data for the site was withheld. In addition to reviewing the results developed, a discussion on the benefits and drawbacks of each approach utilised will also be presented.
Ronalds, R. Rowlands, A. and Zhang, H. (2017) The performance of on-site stormwater detention systems in response to recent advances in hydrologic theory. 13th Hydraulics in Water Engineering Conference. Sydney: Engineers Australia, 2017: 354-362 (link).
Abstract: Recent advances in the theory governing the estimation of hydrographs for urban drainage applications have been adopted by Australian guidelines and public policy. At the core of these changes has been the deviation from previous deterministic empirical formulae toward the analysis of a statistical variation of input parameters that result in multiple hydrograph possibilities for a single design event.
The present study raises concerns regarding the performance of stormwater detention systems for urban drainage applications that have been designed and constructed using the previous ARR1987 methods of deterministic empirical hydrograph generation.
This study considers the potential for the failure of existing detention systems in achieving their stated objectives when these systems are re-assessed using multiple input hydrographs. The effects of multiple input hydrographs are examined by analysing the results of numerical experimentation.
Several case studies with pre-constructed detention systems that have been optimally designed using previous deterministic empirical hydrograph techniques are examined in this experimentation. Using currently adopted ensembles of temporal patterns to assess the potential for variability in hydrograph shape, the case studies are examined to determine the potential for failure, as well as the patterns of failure in the data set.
Opinions and issues regarding the probability and cause of failure to meet objectives using new techniques are presented and discussed, as well as the potential impact on surrounding infrastructure and receiving waterways.
Scorah, M., Stephens, D., Nathan, R. (2019) Benchmarking the selection of probability neutral hydrologic design floods for use in 2D hydraulic models. Australasian Journal of Water Resources https://doi.org/10.1080/13241583.2019.1603334
Abstract: In the 2016 release of Australian Rainfall and Runoff (ARR), variability of flood influencing factors (such as losses and temporal pattern) is now explicitly considered through application of Monte Carlo and ensemble approaches. However, hydrologic models are commonly an input for 2D hydraulic modelling, where computational constraints prevent a large number of floods from being modelled. Practitioners are therefore often required to select ‘representative floods’ for use in hydraulic modelling. When making this selection there is an implicit assumption that the probability of rainfall directly corresponds to the probability of flood levels—that is, the transformation is ‘probability neutral’. This assumption is often untested. The ‘representative flood’ approach was benchmarked against probability neutral estimates of flood depths constructed through Monte Carlo application of a 2D hydraulic model. Hydrographs that were found to be probability neutral with regards to peak outflows from the hydrologic model did not necessarily result in probability neutral estimates of flood depths. This demonstrates the need for running a greater selection of events to avoid generating biassed hydraulic model results.