A new way to geoengineer landscapes using computer-based landform evolution models

Abstract

The design and construction of post-mining landforms is a complex undertaking where any structure requires integration with underlying materials and the surrounding unmined or undisturbed landscape. A common reconstruction design for post-mining landscapes is to have linear hillslopes with drains or runoff diversion structures that are designed for the hillslope length, angle and climate. These landscapes are easy to construct and result in a surface which can be easily traversed by agricultural machinery, while the benches often rely on drainage control structures to manage runoff and resultant erosion. Few mines worldwide have committed to a catchment-based reconstruction approach or that employing geomorphic design. Here, a method for catchment design has employed a simple strategy of an uplifted catchment being allowed to evolve using a computer-based Landscape Evolution Model until the volume matches that of a proposed design. The computer-generated landforms are compared with that of a catchment created using site hydrology and sediment transport conditions (Expert Knowledge) by a recognised design engineer. The results demonstrate that a computer-generated landscape produces sediment output within that of target erosion rates with low gully depths. The design created using Expert Knowledge produces sediment output above background erosion rates as well as having maximum gully depths of up to 2.7 m. Modelling demonstrates that computer-generated designs produce erosion rates which are approximately one-third to half that of the Expert Knowledge design, with a commensurate reduction in maximum gully depth. The computer model-generated catchments also have a more natural appearance with regular curvature and channel definition. A key finding is that landscapes with a series of smaller catchments and a more complex drainage network produce less sediment output.