Accurate prediction of soil erosion is essential for sustainable land management, particularly in agricultural and post-mining landscapes. However, reliable erosion prediction requires robust parameterisation and careful comparison of model outputs with field-based estimates. This study combines field observations with laboratory flume-experiment data to parameterise two landform evolution models (LEMs), SIBERIA and SSSPAM, for simulating soil erosion in the Stanley Jr. catchment (∼8 ha) in southeastern Australia. Both models were run using vegetated and unvegetated parameter sets and evaluated against observed erosion rates derived from long-term environmental tracer ( 137 Cs) measurements and sediment deposition data. Results show that SIBERIA and SSSPAM produced erosion estimates consistent with field measurements, with both models generating variable annual sediment outputs. The flume experiments provided controlled insights into vegetation effects, revealing a notable increase in erosion following removal of above-ground biomass. Despite limitations associated with laboratory-scale parameter derivation, both models demonstrated reliability in low-erosion agricultural environments. Overall, the findings highlight the utility of SIBERIA and SSSPAM for predicting soil erosion and supporting land management and planning. • SIBERIA and SSSPAM accurately predict erosion across lab and catchment scales. • Flume experiments quantify vegetation effects on erosion in controlled conditions. • Models validated against tracer data and sediment basin measurements. • Integration of lab and field data strengthens landscape-scale erosion predictions. • Both LEMs are reliable tools for agricultural and post-mining land management.
Hancock et al. (2026) studied this question.