Comparison between the orographic response of the ECMWF model and the pyrex 1990 data

Abstract

AbstractRecent observations and analyses of the atmospheric flow in the vicinity of the Pyrénées made during PYREX are compared with numerical experiments carried out with the T213L31 European Centre for Medium‐range Weather Forecasts (ECMWF) global model. This study shows that the model simulates a certain proportion of the measured pressure drag, but always underestimates it. In the current model, this deficit is not adequately made up by any of the parametrized subgrid‐scale processes. When the flow is along the range, the pressure‐drag value depends on the synoptic background flow structure. It is close to zero when the flow impinging on the range is baroclinic. It is not zero when the atmosphere is more barotropic (e.g. in the absence of any fronts) and the model misses important lift processes. When the flow is normal to the ridge, the model response to the mountain forcing is essentially ageostrophic. In this case comparison between the vertical profiles of the model stresses and the measured profile suggests that further drag should be applied in the model at low level. This is supported by comparisons of isentropic diagnostics, from the ECMWF model and the French mesoscale model, Peridot. These diagnostics show that the ECMWF model underestimates irreversible low‐level processes, such as the flow separation on the mountain flanks and the associated downstream wake, which can be related to the drag. They also show that low‐level parametrized drags have a realistic impact on the model simulation of the mountain wake. This analysis supports the current developments at the ECMWF of a new subgrid‐scale orographic drag scheme which parametrizes the low‐level impact of mesoscale mountains.