The processes governing horizontal resolution sensitivity in a climate model

One of the questions that climate modellers should address is whether their models have sufficient spatial resolution to resolve the physical processes affecting climate. This study addresses this question using the Hadley Centre climate model, HadAM3 (Hadley Centre Atmospheric climate Model version 3, the climate version of the Met Office's Unified Model). The model is run in AMIP2 (Atmospheric Model Intercomparison Project number 2) mode at four resolutions ranging from N48 (2.5degrees x 3.75degrees) to N144 (0.833degrees x 1.25degrees). The convergence of the model on increasing resolution is evaluated, and the processes leading to resolution sensitivity are investigated in some detail. A parallel set of four dynamical core integrations give an indication of the sensitivity of the dynamics with simple physical parametrization feedback. Increments from individual parametrization schemes during short 'spinup' integrations with full physics are used to diagnose the sensitivity of individual schemes. The dependency of the results on particular details of the model are also investigated to see how general the results are. Model biases are reduced when resolution is increased. In particular, the troposphere warms, the jets shift polewards, mean sea level pressure over the poles falls, and transient vertical velocity and eddy kinetic energy Increase. Some features of the model converge while a number of significant features do not. Non-linearity in both the hydrological cycle and the dynamics play an important role in the lack of convergence.