RORB Version 5: A Tool to Move Beyond ARR87

The 1987 edition of Australian Rainfall and Runoff (ARR87) provides detailed guidance on the calibration and application of runoff-routing models for the purpose of deriving flood estimates of specified annual exceedance probability (AEP). This guidance is based on the 'Design Event' approach in which all parameter values and inputs other than rainfall are treated as fixed values. Careful attention must be given to ensure that the major factors that control the transformation from rainfall burst to flood hydrograph are 'probability-neutral'; that is, that the AEP of the floods is the same as the input rainfall. Considerable research has been undertaken over the past two decades since the guidance presented in ARR87 was formulated. This peer-reviewed work has yielded concepts and design guidance that in many cases provides a more defensible means of deriving probability-neutral estimates of flood hydrographs than is provided in ARR87. The manner in which rainfall bursts, losses, and the temporal and spatial patterns of rainfall are best specified has received a lot of consideration, and additional evidence is now available for the development of regional prediction equations of both losses and model parameters. Importantly, a lot of attention has been given to the stochastic manner in which flood modifying factors interact, and there are now practical joint-probability methods available that avoid the limitations of the Design Event approach. This paper describes the development of a new version of the RORB model that allows practitioners to selectively depart from what is recommended in ARR87 for design flood estimation. The model allows users to optionally take advantage of improved design procedures, including a practical means of using Monte-Carlo simulation to combine losses and temporal patterns in a probability-neutral manner. The rationale for the possible options is discussed, and some practical examples of application of the new model version to design are presented.