Global prediction is shifting towards earth system models where different models (e.g., atmosphere, ocean, wave, sea ice, aerosols, land) are coupled to produce short-range to seasonal forecasts. In NRL's atmospheric global modeling group, basic and applied research is being performed on (1) coupling mechanisms and processes, (2) subseasonal-to-seasonal prediction, (3) model diagnostics, (4) systematic biases in coupled models, and other aspects of coupled modeling. We are interested in prediction in many regions (e.g., polar, tropical, mid-latitudes) and the teleconnections between these regions. Potential research topics include, but are not limited to parameterization development, earth system processes, predictability, ensemble development, data assimilation, and post-processing. Polar atmospheric prediction and diagnostics are a major focus because of the expected increase in Arctic operations. Another focus area is improving subseasonal-to-seasonal tropical cyclone prediction through improving physics parameterizations and post-processing.