Including soil erosion and sediment delivery to surface waters in a high-resolution global hydrological model

Soil erosion and sediment delivery to surface waters impact the human-water cycle in several ways. Eroded soil, along with nutrients, travels from intensive crop and grazing systems to waterbodies, reducing the fertility of agricultural land, degrading the integrity of freshwater ecosystems, and negatively impacting water quality on which the human population and other sectors depend. Furthermore, sedimentation reduces the functional capacity of vital energy and agricultural infrastructure, such as reservoirs and irrigation canals, resulting in reduced productivity and profitability of food and energy production systems. A few hydrologic models have accounted for the effects of soil erosion on water quality by implementing soil erosion models into their simulations. To our knowledge, soil erosion has not yet been included in large-scale hydrologic models. Our research adds an erosion-sediment transport module to the hydrological Community Water Model (CWatM). That is to assess soil erosion on a regional to a global scale and to simulate the concentration of suspended sediments in surface waters. CWatM is a fully-distributed, large-scale, open-source hydrological model. It runs on a daily time step and high resolution of up to 30 arc seconds (approximately 1 km at the equator). The model can account for human activity and management of water systems, including reservoir operations, water demand, and crop-specific irrigation requirements. A global dataset of 5 arc minutes is available for an easy simulation setup at a catchment, region, and global scale. The implementation of soil erosion and sediment delivery from terrestrial sources will rely on the Modified Universal Soil Loss Equation (MUSLE) and the stream network density within each grid cell. Further, simulating instream erosion uses a power law approach. Finally, the routing algorithm would move suspended soil particles downstream. For that purpose, we combine input datasets with CWatM variables, e.g., surface runoff. We apply this module to a case study in Uganda’s share of the Victoria Lake Basin at five arc minute resolution, where high soil erosion rates challenge the ecological integrity of the natural environment, as well as agricultural productivity and water quality. We further discuss the model sensitivity to input parameters’ variation (e.g., the fraction of daily rainfall in the half-hour of highest intensity).