Integration of WBNM into A Continuous Simulation System for Design Flood Estimation

The main approach to design flood estimation in Australia is event-based with the use of flood hydrograph models such as RORB, WBNM and URBS the most common practice for estimating design floods via rainfall based approaches. A characteristic of these models is that the routing of rainfall excess is distributed within the catchment, the amount of routing depending on the amount of sub-division of the catchment. Continuous simulation of losses is now established as an alternative to event-based design. The Continuous Simulation System for Design Flood Estimation (CSS) is currently the main continuous simulation system in use in. The CSS uses the WBMOD flood hydrograph model which routes rainfall excess at a concentrated storage at the outlet of the catchment instead of distributed through the catchment. This paper describes a comparison of the different methods of runoff routing when used in a continuous simulation system. Using the same data and loss system calibration used in a benchmark study for the CSS, the WBMOD was replaced with the Watershed Bounded Network Model (WBNM) and the benchmarking study was repeated. The results show comparisons of the two runoff routing models as well as comparisons of design flood estimates by flood frequency analysis, conventional event-based methods, and by the two routing models. Overall, calibration results for the CSS/WBNM were of a similar accuracy as for the original CSS/WBMOD with calibration of a single parameter (lag k) and simulation via a more realistic routing of flows through catchment subdivision based on drainage patterns. The most obvious differences between systems were the timing of flows and consistently lower design flood estimates for the CSS/WBMN.