Abstract The use of draglines to remove overburden in Queensland opencut mines, results in landscapes that consist of long parallel tertiary overburden spoil-piles that are generally highly saline, dispersive, and highly erodible. The height of these spoil-piles may exceed 50–60 m above the original landscapes and the slopes are at the angle of repose of around 75% or 37°. Legislation and public opinion require that these highly disturbed open-cut post-mining landscapes should be satisfactorily rehabilitated into an approved post-mining land use with acceptable erosion rates. Therefore, these slopes must be reduced before the landscape can be rehabilitated. The most expensive component of the rehabilitation process is the re-shaping and preparation of the overburden to create a suitable landscape for vegetation growth. As soils and overburden varies greatly in their erodibilities, the extent and cost of earthworks can be minimized, and rehabilitation failures avoided, if soil erosion from designed landscapes can be predicted using laboratory-based parameters prior to construction of these landscapes. This paper describes the development of a model for that purpose.

A catchment or landscape erosion model MINErosion 4 was developed by upscaling the existing hillslope model MINErosion 3 (So, et al., 2018) and integrate it with both ESRI ArcGIS 10.3 or QGIS 3.16 (freeware), to predict event based and mean annual erosion rate from a postmining catchment or landscape. MINErosion 3 is a model that can be used to predict event and annual erosion rates from field scale hillslopes using laboratory measured erodibility parameters or routinely measured soil physical and chemical properties, and to derive suitable landscape design parameters (slope gradient, slope length and vegetation cover) that will result in acceptable erosion rates. But it cannot be used to predict the sediment delivery from catchments or landscapes. MINErosion 4 was validated against data collected on three instrumented catchments (up to 0.91 ha in size) on the Curragh mine site in Central Queensland. The agreement between predicted (Y) and measured (X) values were very good with the regression equation of Y = 0.92X and an R2 value of 0.81 for individual storm events, and Y = 1.47X and an R2 value of 0.73 for the average annual soil loss. This is probably the first time that a catchment scale erosion is successfully predicted from laboratory measured erodibility parameters.