Distributed and semi-distributed hydrological modelling for catchments draining to the Great Barrier Reef coastline

ABSTRACT The Great Barrier Reef (GBR) is a globally significant natural wonder, threatened by pollutant loads from eastward-draining rivers. A distributed hydrological model, G2G (Grid-to-Grid), has been configured at a resolution of 0.01° (∼1 km) and 1 hour to simulate these flows. A semi-distributed hydrological model, SWIFT (Short-term Water Information Forecasting Tool), has been set up for 9 catchments at 19 streamflow gauging locations within the same area, as part of the Australian national 7-day Ensemble Streamflow Forecast Service. This provided an important research opportunity to assess the two modelling approaches for different purposes. The performance of both models has been evaluated at the 19 common gauged locations and varies from site to site, but is broadly comparable. The SWIFT is slightly better because it is calibrated to each gauged catchment. In contrast, G2G is configured as a single area-wide model supported by landscape spatial datasets and time-series flow data. G2G is capable of generating outputs for any grid cell, including ungauged locations, whilst SWIFT can only generate outputs at predefined locations on the river network. In conclusion, these catchment-specific and area-wide modelling approaches are compared and their suitability discussed in relation to the modelling purpose.