Evaluation of the Weather Research and Forecasting Model in the Durance Valley Complex Terrain during the KASCADE Field Campaign

Abstract In the winter of 2012/13, the Katabatic Winds and Stability over Cadarache for the Dispersion of Effluents (KASCADE) observational campaign was carried out in southeastern France to characterize the wind and thermodynamic structure of the (stable) planetary boundary layer (PBL). Data were collected with two micrometeorological towers, a sodar, a tethered balloon, and radiosoundings. Here, this dataset is used to evaluate the representation of the boundary layer in the Weather Research and Forecasting (WRF) Model. In general, it is found that diurnal temperature range (DTR) is largely underestimated, there is a strong negative bias in both longwave radiation components, and evapotranspiration is overestimated. An illustrative case is subjected to a thorough model-physics evaluation. First, five PBL parameterization schemes and two land surface schemes are employed. A marginal sensitivity to PBL parameterization is found, and the sophisticated Noah land surface model represents the extremes in skin temperature better than does a more simple thermal diffusion scheme. In a second stage, sensitivity tests for land surface–atmosphere coupling (through parameterization of z 0 h / z 0 m ), initial soil moisture content, and radiation parameterization were performed. Relatively strong surface coupling and low soil moisture content result in a larger sensible heat flux, deeper PBL, and larger DTR. The larger sensible heat flux is not supported by the observations, however. It turns out that, for the selected case, a combination of subsidence and warm-air advection is not accurately simulated, but this inaccuracy cannot fully explain the discrepancies found in the WRF simulations. The results of the sensitivity analysis reiterate the important role of initial soil moisture values.