Harley D. Betts

13 Changes in basin-scale sediment supply and transfer in a rapidly transformed New Zealand landscape

Abstract Society has an ever-increasing need to manage landscapes. To do this effectively requires improved understanding of the way landscapes behave, and the controls on that behaviour. This is certainly the case where sustainable resource use and hazard mitigation involve the management of the generation, transport, and storage of sediment. Landscapes are complex systems, consisting of a mosaic of landforms. At the broad regional level these landforms are arranged in characteristic patterns, reflecting environmental conditions and associated processes. At the catchment level, assemblages of landforms have a unique configuration, forming an interactive functioning system through which water and sediment are passed. It is this unique, catchment-based assemblage that we seek to manage. Controls on the way landscapes behave are numerous and operate at a variety of scales both spatial and temporal. The way these controls interact on a complex and unique arrangement of landforms is difficult to predict. The East Coast of the North Island of New Zealand is a dynamic landscape. High natural erosion rates have been augmented by recent and rapid anthropogenic activity. Several studies in the Waiapu catchment, involving a range of spatial and temporal scales, are used here to illustrate the impact of natural and anthropogenic controls on basin-scale sediment supply and transfer. In this landscape, the use of vegetation, specifically targeted reforestation, is the most effective method of sediment management. This will be enhanced by improved understanding of stability thresholds and hill slope–channel connectivity.

Development of a landslide component for a sediment budget model

Most erosion models focus on overland-flow erosion with fewer incorporating landslide erosion although it is common on hillslopes. Landslide models are typically dynamic, spatially distributed simulations with large data requirements for parameterisation and are often computationally intensive. The Australian SedNet model represents a middle ground between process-based and empirical models and is being modified for New Zealand conditions by incorporating shallow landsliding.We describe a method for implementing a model within SedNetNZ to provide the long-term annual average sediment contribution from shallow landsliding and its spatial distribution. The mass of soil eroded over a defined period is calculated from the landslide probability for each slope class, slope class area, failure depth, soil bulk density, and sediment delivery ratio. Landslide probability is derived from mapping a time series of landslides intersected with DEM-derived slope. The conceptual approach and methodology for parameterisation are suitable for landslide modelling where rainfall-triggered shallow landslides occur. A method for modelling long-term shallow landsliding within a sediment budget model.Time series of historical landsliding mapped from aerial photography.Landslide-slope relationships derived from landslide distribution and slope.Landslide-slope relationships used to spatially model sediment generation.Applicable to any landscape subject to rain-triggered shallow landsliding.