Alleviation of stationary biases in a GCM through a mountain drag parameterization scheme and a simple representation of mountain lift forces

The problem of the representation of the orientation of mountain forces in a GCM is examined. First a series of winter simulations is presented with the Laboratoire de Meteorologie Dynamique GCM where the model ground is flat and where the mountains are represented by applying forces to the flow that are either strictly opposite to the local wind (i.e., a drag) or strictly perpendicular to it (i.e., a lift). These experiments show that the drag forces improve the zonal wind but do not improve the steady planetary wave everywhere, while the lift forces scarcely affect the zonal wind but force a steady planetary wave that is everywhere realistic. The drag force is formulated following the realistic subgrid-scale orographic (SSO) scheme, developed recently at ECMWF. The lift force essentially enhances vortex stretching over large-scale mountains. To address issues concerning the representation of SSO in GCMs, experiments with mean orography, the ECMWF SSO scheme, and enhanced lift forces over mountain massifs are compared to experiments with mean orography only. They confirm the results of the 'no mountain' runs: the SSO scheme improves the zonal wind and the enhanced lift forces improve the steady planetary wave. Although these results suggest that the recent SSO schemes that are validated against data, and that significantly decelerate the low-level flow, are beneficial to GCMs, they also raise the question of the representation of the orientation of mountain forces that should be included in these schemes. For this purpose, the lift representation proposed should be considered carefully in an operational context, because it is only linked to the problem of SSO parameterization by qualitative arguments.