The Interface Problem in Model Coupling: Examples from Atmospheric Science

Examples drawn from atmospheric sciences demonstrate that models focus on internal consistency. This optimization, while necessary, prevents a focus on interface adaptation. Model validation primarily assesses the internal model behavior, rather than the adaptation and usability of its interface to the output. With the development of solutions from computer science to overcome issues of data incompatibility, the conceptual issues of model interface coverage gain importance. Covering processes along interfaces may become the major challenge of modeling in the future. The interfaces should in part be regarded problems of communication between scientists, and thus be attributed a sociological approach. The "middle ground" could at the same time describe an ill-defined interface calling for better description, but also reflect a challenge to overcoming scientists´ disciplinary divergence.