| description abstract | The authors use a 1998 workshop titled "Observations, Experiments, and Large?Eddy Simulation" as a springboard to begin a dialogue on the philosophy of simulation as well as to examine the relationship of large eddy simulation (LES) of geophysical flows to both observations and experiments. LES is shown to be perhaps the simplest representative of a broad class of activity in the geosciences, wherein the aggregated properties of fluids are solved for using approximate, or conjectural equation sets. To distinguish this type of activity from direct fluid simulation, the terms pseudofluid and pseudofluid simulation are introduced. Both direct and pseudofluid simulation introduce methodological changes into the science as they propose to provide synthetic, yet controlled, descriptions of phenomena that can then be used to help shape ideas regarding the behavior of real fluids. In this sense they differ from more traditional theoretical activities, whose goal is to provide better/simpler explanations of observed phenomena. However, because pseudofluids, by their very nature, demand testing, they supplant neither observations nor experiments. Instead they define additional opportunities and challenges for these well?established scientific methodologies. Such challenges and opportunities primarily manifest themselves as tests, which are categorized into two types: (i) tests that attempt to justify the method a priori and (ii) tests of hypotheses that are derived from the method. LES is shown to be particularly amenable to both types of tests whether they be implemented using observations or experiments. Moreover, the recent developments in laboratory and remote sensing technologies are shown to provide exciting opportunities for realizing such tests. Last, efforts to better understand LES will have peripheral benefits, both because LES shares common features with, and because LES is increasingly used as a tool to further develop, other types of pseudofluids in the geosciences. For these reasons institutional initiatives to develop symbiotic relationships between observations, experiments, and LES would be timely. | |
