Abstract : 2E.13
Impact of horizontal diffusion, vertical resolution, microphysics, radiation and cloudiness parametrization schemes on fog forecast in valleys
Croatian meteorological and hydrological service
Low quality operational forecasts (using Aladin model) of 2 m temperature and low cloudiness in a stable atmosphere with low-level inversion, low stratus and fog have encouraged testing of different horizontal diffusion, radiation and cloudiness schemes available in Aladin. A simple microphysics scheme, with prognostic cloud water and ice, as well as rain and snow, that has been recently introduced in the model provides further insight into the problem.
A study of a synoptic case marked by a strong temperature inversion, low cloudiness and fog in a wide valley in the inland part of Croatia, that lasted for several days showed that even small changes, like different cloud overlap assumptions, can bring significant improvement in the forecast of low cloudiness and, consequently, the 2 m temperature diurnal pattern. Surprisingly, more sophisticated radiation schemes do not always bring an improvement. Microphysics has a beneficial impact, but has less impact on the low cloudiness and fog forecast than the cloudiness scheme.
The correct forecast of fog and low stratus in narrow Alpine valleys is highly affected by the horizontal diffusion scheme used in the model. Common 4th order numerical horizontal diffusion is applied on model levels that follow orography, thus it is not purely horizontal and it is not physical. The significance of the physical horizontal diffusion increases with horizontal resolution as model levels become more tilted close to mountain areas and the horizontal mixing often occurs between “the valley” and “the mountain top”.
The Semi-Lagrangian horizontal diffusion (SLHD) is based on the control of the degree of interpolation needed for the semi-Lagrangian advection scheme that depends on the local flow. In the case of fog in an anticyclone, use of SLHD increases the amount of fog in Alpine valleys.