The use of limited-area models (LAMs), with their enhanced spatial and temporal resolution, is motived by the desire to more completely resolve the nonlinear interactions among different scales of motion. However, the development of smaller-scale features is intimately related to the larger scale environment in which they are embedded. The largest scale motions, especially those extending beyond the length of the LAM domain, are forced through the lateral boundary condition (LBC) and are typically assumed to be perfect.

Experiments using global model ensemble systems demonstrate the importance of accounting for uncertainty in atmospheric analysis and forecasting. Such experiments motivate the development of new LAM strategies that acknowledge and utilize the range of equally probable LBCs provided by global ensemble systems. In a traditional configuration, model differences and small scale features growing upscale within the LAM domain cause the solution to become inconsistent with the low wavenumber forcing imposed at the boundary. The strategy being developed in this work will allow dynamic testing and selection from among the ensemble of LBCs, thereby providing consistency between features evolving at scales common to both domains. While improving validation statistics for individual LAM forecasts, it is hoped the results of this work will also improve the dispersion characteristics for short-range LAM ensemble forecasts.