Rehabilitated mine landforms need to remain stable long after reshaping is complete. The challenge is that erosion does not always begin at one simple, universal threshold.
A central Queensland case study co-authored by our Engineer Mercy Atukunda compared five prominent gullies with predicted soil loss using the Revised Universal Soil Loss Equation (RUSLE), supported by aerial imagery, LiDAR and GIS analysis. The results showed that the soil-loss rate associated with gully initiation varied as slope length and gradient changed. This points to a more site-specific way of designing and managing rehabilitated landforms.
What was the purpose of the study?
The study looked at the relationship between where gullies actually formed on a reshaped mine landform and the soil-loss rates predicted by RUSLE. The aim was to test whether observed erosion could be used to improve confidence in landform design parameters.
How was the model checked against what was happening on the ground?
The team used aerial imagery, LiDAR and GIS analysis to examine five prominent gullies that developed about three years after reshaping. Their slope lengths and gradients were then compared with RUSLE-predicted soil-loss rates. RUSLE is a widely used model for estimating long-term average soil loss. It considers factors such as rainfall, soil erodibility, slope length and steepness, surface cover and erosion-control practices.
What did the five gullies reveal, and why does it matter for landform design?
The five gullies formed across slope lengths of about 50 to 114 metres, with predicted annual soil-loss rates ranging from around 140 to 212 tonnes per hectare per year. Importantly, there was no single erosion threshold at which gullies began to form. This shows that slope length and gradient need to be considered together when designing rehabilitated landforms, rather than applying one generalised erosion threshold across an entire site.
How can these findings improve rehabilitation design in practice?
The findings support tailoring erosion controls, such as contour drain spacing, to the conditions of individual slopes rather than using the same approach everywhere. More broadly, they show the value of testing models against real site behaviour. Combining modelling with aerial imagery, LiDAR, GIS and field observations creates a practical feedback loop for refining designs and improving long-term landform performance.
