Estimating soil erodibility for the RUSLE with rainfall simulation in central Queensland, Australia

Context Land use change in central Queensland has led to increased sediment and nutrient loads in the Great Barrier Reef (GBR) lagoon, impacting water quality. Accurate soil loss prediction is essential for managing these impacts, necessitating improved erodibility estimates for grazing lands. Aims The objective of this study was to advance the prediction of fine sediment exports to the GBR lagoon from hillslope sources. It specifically aims to enhance hillslope erosion modelling by refining the Revised Universal Soil Loss Equation (RUSLE) soil erodibility estimates (K-factors) for grazed lands. Method An integrative approach was employed, combining the empirical RUSLE with the process-based Water Erosion Prediction Project (WEPP). This study leverages observed soil loss data from simulated rainfall experiments to calibrate WEPP and produce an annual average soil loss to substitute into RUSLE. Prevalent soil types in the Fitzroy Basin were studied. Key results The WEPP was calibrated with observations and shown to be able to simulate observed sediment loss with good performance indicator values (R2 = 0.9, PBIAS = 1.9%, and NSE = 0.87). The study identified that the existing RUSLE nomograph-derived K-factors would overestimate soil erodibility by up to 63% for the soils evaluated. Traditional methods predict higher erodibility; however, results from this study classify these soils as having low to moderate erodibility, attributed to local soil consolidation and limited erosion detachment processes typical in grazed areas. Conclusions Findings suggest that the conventional RUSLE erodibility nomograph does not adequately reflect the erodibility of consolidated or highly aggregating grazing soils found in the GBR catchments, leading to overestimated K-factor values and, subsequently, overestimated contribution of hillslope erosion to the sediment budget for the GBR catchments. Implications This research contributes to delivering a cost-effective, measurement-based method for K-factor estimation to improve hillslope erosion prediction for grazing lands in central Queensland.