Combining idealized dynamical core test cases with regionally refined meshes (RRM) enables us to investigate nonhydrostatic effects in multiscale simulations independent of the details of subgrid-scale parameterization packages. By comparing small-planet and full-planet RRM simulations, we investigate the feasibility of simulating storm-scale phenomena at 1km or smaller in the ACME global climate model. Further, we are interested in determining at what scales nonhydrostatic phenomena become important. Although analytic scaling arguments predict a 10km threshold, the effective cut-off resolution may be both model and problem dependent. By tuning the small-planet radius, we are able to identify the emergence of nonhydrostatic effects and quantify the hydrostatic approximation error as a function of scale.