On the numerical asymmetry in calculating Coriolis terms through the splitting method in a mesoscale model

The asymmetry in calculating the Coriolis terms by Marchuk's splitting method is discussed. It was found that although some asymmetry is required for linear stability, this asymmetry should be kept to a minimum. Otherwise, for instance, if the diffusion terms are calculated in between the Coriolis terms in the u and v equations of motion, one gets cross-isobaric low-level winds, which are different depending on the geostrophic wind direction. This is the first study to show that the required asymmetry in the splitting can cause a considerable error in the boundary-layer winds of atmospheric models if not handled in a specific way.