Assessing the effectiveness of erosion predictions from a landform evolution model with field observations from a rehabilitated landform

Landform evolution models (LEMs) can assess the effectiveness of landform designs applied to a rehabilitated landform by predicting where erosion may occur. Here, we compare erosion predictions produced by the CAESAR-Lisflood LEM with field observations from the rehabilitated Pit 1 landform of the Ranger uranium mine, based in the Northern Territory, Australia (Figure 1). The ability to predict long-term erosion at a catchment scale is important, as the final rehabilitated Ranger landform is required to physically isolate buried tailings material for at least 10 000 years. CAESAR-Lisflood has been extensively tested and calibrated using data collected from erosion plots (0.09 ha) on the Ranger trial landform (TLF) for multiple years (Lowry et al, 2020; Saynor, Lowry and Boyden, 2019). However, until recently, it was impossible to compare model predictions with field observations for areas larger than the TLF erosion plots. We can now compare model predictions of Life of Mine Conference 2023 | Brisbane, Australia | 2–4 August 2023 116 gully development with field observations from the newly rehabilitated, 40 ha surface of Pit 1 over two contiguous one-year periods: 2020–2021 and 2021–2022. The aim of this study was to assess whether the CAESAR-Lisflood model can successfully predict gully erosion at a larger spatial scale.