Environmental Modeling Research at the University of North Carolina at Chapel Hill: Final Report

Mechanistic mathematical models of environmental systems are used routinely to assess our understanding of the operative complex processes in nature. As our understanding matures, the complexity of these models increases and so too does the effort required to construct such models. This effort can be person years in some cases, and changes in model formulations or methods frequently leads to the need to either perform major revisions of existing codes or to abandon an existing code and recode the majority of the simulator. This project was intended to be a proof of concept approach aimed at developing a problem solving environment for the development of environmental models. The domain of focus was fluid flow and species transport in subsurface, porous medium systems. An approach was developed in which a mathematical model formulation was specified in LaTeX and this text document was processed, or compiled, multiple times to ultimately result in a computational simulator or model. The DOE developed Common Component Architecture paradigm was leveraged to implement solvers for reactions, integrators, algorithms, and discretization methods. A set of test problems was solved and the overall conclusion of the project is that a problem solving environment to support environmental modeling is certainly within scientific reach at this time