Spatial Modelling of Floodplain Inundation Combining Satellite Imagery and Elevation Models

The environmental benefits of water flowing into wetlands and over floodplains depend on the extent, depth, duration and frequency of the inundation. Along the River Murray in South-Eastern Australia, the health of forests and the breeding patterns of birds depend directly on the patterns of wetting caused by flooding events. This wetting is a major contributor to the recharge of ground waters. Therefore, river managers need to predict the dynamics of wetting when they plan environmental flows so they can optimise the environmental benefits to the riparian ecosystem. This paper describes the development of a River Murray floodplain inundation model for river managers to predict the impacts of flow regimes. This paper contributes a method for modelling the extent and depth of flooding in a complex river system. First, we investigate how existing methods for modelling flood inundations apply to the River Murray. Noting the significant drawbacks of existing models, we describe a new technique for predicting the daily depth of water on floodplains from high resolution digital elevation data and satellite imagery. Finally, we describe the incorporation of this data into floodplain modelling software, namely the River Murray Floodplain Inundation Model (RiM-FIM). There are many existing methods for modelling the flow of water in floodplains. Hydrologists are familiar with equations governing river flow and with software products such as one dimensional (MIKE11 and HEC-RAS) and two dimensional (MIKE21) hydrodynamic models. However, with very complex river systems such as the River Murray, parameterising these models is expensive. Hydrodynamic models require river cross-sections, accurate high resolution digital elevation models (DEMs) and extensive calibration. Existing hydrodynamic approaches for river modelling are expensive to apply to the River Murray. Therefore, we have developed a technique for predicting the depth of water on floodplains from high-resolution digital elevation data and satellite imagery. The technique involves combining flood masks from Landsat imagery with a very high resolution DEM. The result is a set of very high resolution flood maps for river flows observed at gauging stations. The flood maps in Figure 1 show improvements adding elevation data provides to satellite imagery. RiM-FIM incorporates this data with historical gauging station flows to produce a temporal model of floodplain inundation. The model enables users to visualise and analyse the inundation of floodplains along the River Murray.