Modelling Soil Erosion on a Global Scale – An Earth System Model Approach

Abstract:

The current Dynamic Global Vegetation Models (DGVMs) do not treat soils as a dynamical system and lack the capability to account for soil erosion processes (detachment, deposition and sediment transport), which limits their usage for investigating soil carbon dynamics and the global carbon cycle. We address this problem by developing a soil erosion module for application in the Max Planck Earth System Model (MPI-ESM), which could represent the different soil erosion processes at a global scale. This module computes soil detachment first, which is then integrated in a simple sediment mass-balance equation to estimate sediment deposition and transport. Soil detachment is modelled by implementing the Revised Universal Soil Loss Equation model, RUSLE. For this, we use scaled global datasets of topography, precipitation, soil and land-use as input data. We also perform a sensitivity study on the topography and precipitation factors in the RUSLE model, and show that there is a large uncertainty involved when applying these RUSLE factors at a global scale. The sediment mass-balance equation is imbedded in a cascade grid structure that represents sediment flow and considers soil detachment from the RUSLE model and a deposition term to model alluvial and colluvial sediment storages and fluxes. As a first step towards a large scale implementation, the complete soil erosion module is applied on a 10km spatial resolution on the Rhine river catchment and is validated with existing sediment data. We find that the soil erosion module operates well at this spatial scale for the Rhine catchment and is thus promising for the application in MPI-ESM. However, the challenge remains to make the model applicable on a global scale. Hence, we present the different challenges of applying the module for other catchments that are physically different from the Rhine catchment, and show the perspectives of using the module as a basis for a global carbon mass balance model.