Water Resources Research Act Program

Details for Project ID 2016TN119B

Development of a Robust Model for Cross-Scale Prediction of Flow and Sediment Transport

Institute: Tennessee
Year Established: 2016 Start Date: 2016-03-01 End Date: 2017-02-28
Total Federal Funds: $5,000 Total Non-Federal Funds: $2,550

Principal Investigators: Benjamin Abban, Thanos Papanicolaou

Project Summary: Current research objectives are to develop at state-of-the-art tool capable of simulating the transport of water and sediment from the plot scale to the watershed scale, while capturing important feedback effects across the scales. Such a tool is needed for effectively designing and monitoring practices for mitigating the impact of intensive management of landscapes by humans. Nearing the end of my PhD (expected in July, 2016), I have so far developed a coupled numerical model that links upland processes with in-stream processes in a dynamic fashion, and in addition is able to fully quantify uncertainty related to model predictions. The tool has been developed using several programming languages (C, C++, Fortran), statistical tools (OpenBUGS, R) and existing established models (3ST1D, WEPP). Several components of the tool have been presented in conferences, published as part of conference proceedings, or are in the process of being presented in a manuscript for peer-reviewed journals (see references below). The tool is a prime candidate for inclusion in the Community Surface Dynamics Modeling System (CSDMS), which is a collection of software for the earth science community to advance our understanding of earth surface processes. The tool, however, has one major bottleneck despite its utility. It is computationally demanding and so requires significant amount of time to perform a simulation. Thus, to further enhance the utility of the model and facilitate its ready adoption by the community, I am proposing to complete the following additional tasks for which I would like to be funded: Develop a parallel version of the tool that can run on everyday graphical processing units (GPUs) to considerably reduce computational time. Since most new desktops now are configured with GPUs, this enhancement should be well-received by the community. Present this novel work (in its entirety) to the earth science community at the upcoming European Geosciences Union (EGU) General Assembly in 2016, in hopes of getting valuable feedback from the entire community that can be used to further improve the tool and enhance our knowledge of cross-scale landscape processes. Publish the completed model in a peer-reviewed journal paper.