Some implications of simulating the Earth System

The capability to simulate, rather than model the Earth System, implies the use of solvers applied to equations that generalize to other situations.  For instance, km-scale storm resolving models solve the same equations as used in LES of the marine boundary layer at much higher resolution, the dynamics of a buoyant thermal, or the interaction of flow with topographic features.  By limiting the use of models (parameterization) to the non-fluid component, e.g., photons or droplets, rather than a part of the flow that is artificially severed from the rest, e.g., convection, scale separation can be better enforced, and the ensuant models can be put on a better theoretical footing.  This approach results in more physical models grounded in assumptions that are countably small, often separable, and more directly comparable to observations.  That creates new opportunities to test the ability of the models to represent constituent processes within the Earth system, and a new basis for developing physical understanding. This leads to new scientific opportunities which we highlight by reviewing the diversity of configurations being applied, the problems they are solving, and the benefits being derived from their compatibility with advanced observations — including those from an emerging and a promised new generation of active satellite remote sensing.