Investigation of dominant hydrological processes in a tropical catchment in a monsoonal climate via the downward approach

This study explores the dominant processes that may be responsible for the observed streamflow response in Seventeen Mile Creek, a tropical catchment located in a monsoonal climate in Northern Territory, Australia. The hydrology of this vast region of Australia is little understood due to the low level of information and gauging that is available. Any insights that can be gained from the few well gauged catchments that exist can be valuable for predictions and water resource assessments in other poorly gauged or ungauged catchments in the region. To this end, the available rainfall and runoff data from Seventeen Mile Creek catchment are analyzed through the systematic and progressive development and testing of rainfall-runoff models of increasing complexity, by following the “downward” or “top-down” approach. At the end a multiple bucket model (4 buckets in parallel) is developed. Modelling results suggest that the catchment’s soils and the landscape in general have a high storage capacity, generating a significant fraction of delayed runoff, whereas saturation excess overland flow occurs only after heavy rainfall events. The sensitivity analyses carried out with the model with regard to soil depth and temporal rainfall variability reveal that total runoff from the catchment is more sensitive to rainfall variations than to soil depth variations, whereas the partitioning into individual components of runoff appears to be more influenced by soil depth variations. The catchment exhibits considerable inter-annual variability in runoff volumes and the greatest determinant of this variability turns out to be the seasonality of the climate, the timing of the wet season, and temporal patterns of the rainfall. The water balance is also affected by the underlying geology, nature of the soils and the landforms, and the type, density and dynamics of vegetation, although, information pertaining to these is lacking