Accounting for the Role of Landscape Heterogeneity in Physically-based Models: Linking Upland and In-stream Processes

The sustainability of croplands and aquatic habitats can only be achieved if effective soil erosion management practices are put in place. To quantify accurately the success of these practices across different scales, a more complete understanding is needed of the processes that lead to soil erosion, as well as the transport of the eroded sediments to and through the stream channel and the watershed outlet. Whilst reliable models have been developed separately for simulating erosion processes on hillslopes and in streams, there is still no model that adequately links the hillslope and in-stream processes. This limits the ability of watershed managers to study, implement, and evaluate erosion protection measures at scales larger than the field. A requirement for linking hillslope and in-stream process-based models is the ability to simulate accurately the conditions that occur in nature. However, for hillslope models, they often assume a homogenous hillslope for the purposes of flow and sediment routing even though they require spatially and temporally heterogeneous data as inputs. The models aggregate certain physical properties of the hillslope in order to simplify the computational workload. This aggregation may introduce errors in simulated flow rates and depths, as well as the sediment yield at the outlet of a hillslope. Thus, on-going field experiments have been designed to improve our understanding of the role of heterogeneity on hillslope erosion for implementing into a recently coupled hillslope/ in-stream erosion model, namely the Watershed Erosion Prediction Project - Steep Stream Sediment Transport 1D model (WEPP-3ST1D). The experiments will identify the important variables that should be simulated in order to account for different types of heterogeneity in coupled process-based models. The overarching goal of the experiments is to outline a comprehensive methodology, by which new and existing models, such as WEPP-3ST1D, may incorporate heterogeneity, thereby providing a better link between upland and in-stream processes. This will lead to more reliable flow and sediment predictions to evaluate the effectiveness of soil erosion management practices across different scales.