Comparison between two single-column models designed for short-term fog and low-clouds forecasting

Resumen

There is a large impact of fog and low clouds in the field of airborne transportation. As a consequence, there is an increasing demand of precise forecasts in order to mitigate the negative effects of visibility reduction. Nevertheless, the production of these forecasts constitutes an unsolved issue yet. In several places, one-dimensional (single-column) models have been tested for local forecasts. They provide many insights into the complex physical mechanisms that are relevant in the fog or low clouds formation. Nevertheless, any influence related with surface or atmospheric horizontal unhomogeneity remains, in most cases, out of control. In 2005, an intercomparison experiment among different single-column models in the field of fog and low clouds forecast was carried out. Since the results were useful but inconclusive, it was agreed to undertake a massive and systematic comparison between H1D and COBEL-ISBA. During the intercomparison period, the models have been run for Paris-Charles de Gaulle International Airport. The most significant difference between the models is found in the nocturnal cooling rate, which is -on average- stronger in H1D. The reason could lie in the fact that H1D predicts less cloud coverage than COBEL-ISBA. The verification results show that the initialization is an important aspect for very short-term forecast and COBEL-ISBA seems to perform better for this point. They also demonstrate that the treatment of horizontal heterogeneities is important for longer forecast scopes, and H1D seems to perform better for this aspect.