RDycore-sediment

Résumé fourni par la source

This archive contains the exact RDycore-sediment source code used in the study describing RDycore-sediment. RDycore-sediment is an extension of RDycore, a finite-volume hydrologic and hydrodynamic model for solving shallow-water flow over unstructured meshes. This version adds suspended sediment transport and bed exchange processes, including multi-class sediment advection, erosion, deposition, active-layer/substrate bed exchange, external soil erosion input, sediment boundary conditions, and diagnostic sediment flux output. These capabilities were implemented to support process-based simulation of sediment mobilization, transport, deposition, and export during flood events. The simulation configurations used in the study are included under: RDycore-sediment/docs/user/example-cases Key cases include the sediment dam-break benchmark, lake-at-rest/c-property test, manufactured-solution convergence test, and Houston Hurricane Harvey simulations with ocean and fixed downstream boundary conditions: - dam_break_sed/: passive dam-break sediment transport test, - c_property_sed/: lake-at-rest/c-property test, - dam_break_sed_new/: dam-break test with erosion and deposition,- mms/: manufactured-solution convergence test, - harvey-flooding/ocean_BC/: Houston Hurricane Harvey sediment simulation with ocean downstream boundary condition, - harvey-flooding/fixed_BC/: Houston Hurricane Harvey sediment simulation with fixed downstream boundary condition, - harvey-flooding/hydro_ocean_BC/: hydrodynamic-only Harvey simulation used to check that sediment transport does not feed back on hydrodynamics. These cases provide the model configurations used for numerical verification, model intercomparison, and the flood-scale sediment transport application. The production Houston Harvey simulations were run on the DOE-supported Perlmutter system at the National Energy Research Scientific Computing Center (NERSC). RDycore-sediment can be compiled and run on other HPC systems, but full production reruns may require large mesh, forcing, and boundary-condition files, as well as machine-specific build and job-submission settings. Users planning to rerun the full Harvey simulations outside Perlmutter should contact the authors for guidance on obtaining and staging the large input files and adapting the build/run scripts to their computing environment.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.