The coupling of the dynamics and physical parameterisations in the next UK Met Office atmospheric dynamical core (ENDGame): idealized experiments

The coupling of the dynamics and physical parameterisations in the next Met Office atmospheric dynamical core (ENDGame) will be presented together with results from idealised Held-Suarez and Frierson-Held-Zurita-Gotor experiments.

ENDGame is an iterative semi-implicit semi-Lagrangian dynamical core which permits the use of relatively large time steps (reaching up to 30 minutes for climate simulations). With such time steps it is believed that it is important to couple the dynamics and physics parameterisations accurately to avoid the numerical error of the coupling dominating the overall model error. In the current, non-iterative model, the so-called slow physics are coupled in parallel whilst the fast physics are coupled sequentially. Iterating the time step, as the new model does, presents a number of alternatives for generalising this approach. Two possibilities that have been considered will be presented.

These approaches were initially tested using the dry Held-Suarez test. However, no significant differences between the various coupling approaches were found. One reason might be the limited time variability and forcing on suitable timescales. It seems likely that the inclusion of moisture and the strong forcing provided by latent heat release might be needed to distinguish between the different coupling approaches. Therefore the grey radiation aquaplanet moist GCM test introduced by Frierson, Held and Zurita-Gotor has been implemented in ENDGame. Results from this test will be presented using the two coupling strategies.