Thursday, 28 June 2007: 9:15 AM
Summit A (The Yarrow Resort Hotel and Conference Center)
The advance in operational forecast of fog has been slow although numerical weather prediction (NWP) models have been upgraded for several generations. Current numerical prediction of fog is conducted locally by running an one-D, detailed in physics, and very high resolution fog model at particular locations forced by a mesoscale background model. For operational forecast on large domain however, such approach requires significant computing resources beyond the capacities of most weather forecast centers. In this paper, two alternative realistic solutions are presented. The first solution is based on the diagnosis of cloud base and top in post processor. This method has been applied in NCEP's Short Range Ensemble System (SREF). The drawback is that it only predicts occurrence of fog but no fog liquid water content (LWC) information which is required for computation of visibility range. The second solution was suggested based on an asymptotic formulation of fog LWC developed recently at NCEP and could be implemented in model post processor. This approach can predict not only fog occurrence conditions at a grid point but also vertical distribution of fog LWC at that point.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner