Different alternatives to high resolution can be applied to reduce our computational needs and understanding. For example, lower resolution models allow scientists to study ensembe integration, parameter sensitivity, and perform bifurcation analysis. Those alternatives include exploiting idealized models, nesting and adaptive mesh refinements, and novel approaches to parametrizations of subgrid processes. However, in order to apply one or more of these tools, better understanding of weather and climate systems is required. First, if we want to have an alternative to a high-resolution run and we aim to resolve multiscale phenomena then we should decide on whether one of the tools listed above is preferrable to the others. Second, we need to know more about each of the tools. For example, for both stochastic and deterministic parametrizations, we should define what kind of particular parametrization to use and how to choose its parameters. For an idealized model, we should be careful about oversimplifying the model.
However, there are features in weather and climate models that cannot be resolved neither by high-resolution runs nor by using the alternatives to high-resolution model runs. The unresolved effects include various types of diffusion, a systematic error, and a model error.

  • No labels