Investigation of the difference in modelling approaches between 1D and 2D flood models

Publication Type:
Thesis
Issue Date:
2009
Full metadata record
Files in This Item:
Filename Description Size
01Front.pdf7.81 MB
Adobe PDF
02Whole.pdf178.5 MB
Adobe PDF
NO FULL TEXT AVAILABLE. This thesis contains 3rd party copyright material. ----- The average annual cost of flooding in Australia is $314 million (BTE, 2001) (or $318 million in 2008 dollars). Approximately 100 flood studies, floodplain management studies and plans are currently being undertaken in Australia to seek to reduce the potential flood risk to residents and properties in flood-affected areas. Consequently, a great body of knowledge and experience in flood modelling practices has been acquired, ranging from the use of one-dimensional (ID) steady-state models through to two-dimensional (2D) finite element hydrodynamic models. Commonly used hydraulic modelling programs are: HEC-RAS (ID), MIKE-11 (ID), RMA-2 (2D finite element), TUFLOW (1D/2D finite difference), MIKE-FLOOD (1D/2D finite difference) and SOBEK (1D/2D finite difference). It is recognised that there are a number of advantages in using 2D modelling. These include: ■ the ability of the model to determine actual flow paths across the floodplain based on the terrain of the study area; ■ simulation of the hydraulic interactions between the channel and floodplain, including momentum exchange, in the horizontal plane; ■ water levels and flow velocities are available at every model grid point; and, ■ the ability to employ fine model resolution near critical floodplain features. However, 2D modelling has significant disadvantages when compared with ID modelling, such as: ■ potential loss of channel definition; ■ more advanced data requirements; ■ difficulty in data integration; ■ greater expense both in terms of initial output and gathering data; and, ■ longer computation times. This thesis critically appraises ID and 2D hydraulic modelling techniques based on a quantitative comparison of the modelling results for two case studies in New South Wales, Australia. These case studies were selected to represent two different types of channel and floodplain systems with distinctive hydraulic features. The first case study involves one-dimensional HEC-RAS and MIKE-11 and two-dimensional RMA-2 modelling of South Creek and its tributaries in Western Sydney. South Creek is a significant tributary of the Hawkesbury-Nepean system and drains a 570knV catchment. The South Creek system comprises a dendritic network of creek channels, with a distinct hierarchical structure between the main South Creek channel and its major tributaries (Ropes, Kemps and Badgerys Creeks) and sub-tributaries. The creek channels only have the conveyance capacity to carry minor flows. Accordingly, the majority of the flow at the peak of major flood events within the catchment is conveyed by overbank areas. The second case study involves one-dimensional HEC-RAS modelling and two-dimensional RMA-2 modelling of the Edward River in the vicinity of the town of Moulamein in western NSW. The study also includes Billabong Creek, which joins the Edward River at Moulamein. This study area is characterised by vast, flat floodplain areas and large river/creek channels that have the capacity to carry up to 20 year ARI flows along some reaches. For each case study, the reliability of the ID and 2D modelling approaches is assessed to determine their applicability to each system. As a result, the thesis provides practical guidance for modellers on the suitability of ID and 2D modelling for a number of physical channel and floodplain characteristics.
Please use this identifier to cite or link to this item: