Distributed empirical algorithms to estimate catchment scale sediment connectivity and yield in a subtropical region

Abstract

The intensity of soil loss and sediment delivery, representing hydrologic and geomorphic processes within a catchment, accelerates with rapid changes in land cover and rainfall events. An underlying component of sustainable management of water resources is an understanding of spatial and temporal variability and the adverse influences of regional parameters involved in generating sediment following widespread changes in land cover. A calibrated algorithm of soil loss coupled with a sediment delivery ratio (SDR) was applied in raster data layers to improve the capability of a combined model to estimate annual variability in sediment yields related to changes in vegetation cover identified by analyses of SPOT imagery. Four catchments in Kangaroo River State forest were assessed for annual changes in sediment yields. Two catchments were selectively logged in 2007, while the two other sites remained undisturbed. Results of SDR estimates indicated that only a small proportion of total eroded sediment from hillslopes is transported to catchment outlets. Larger SDR values were estimated in regions close to catchment outlets, and the SDR reduced sharply on hillslopes further than 200-300 m from these areas. Estimated sediment yield increased by up to 30% two years after land cover change (logging) in 2009 when more storm events were recorded, despite the moderate density of vegetation cover in 2009 having almost recovered to its initial pre-logging (2005) condition. Rainfall had the most significant influence on streamflow and sediment delivery in all catchments, with steeply sloping areas contributing large amounts of sediment during moderate and high rainfall years in 2007 and 2009. It is concluded that the current scenario of single-tree selection logging utilized in the study area is an acceptable and environmentally sound land management strategy for preservation of soil and water resources.