A New Visibility Parameterization for Warm-Fog Applications in Numerical Weather Prediction Models

Abstract The objective of this work is to suggest a new warm-fog visibility parameterization scheme for numerical weather prediction (NWP) models. In situ observations collected during the Radiation and Aerosol Cloud Experiment, representing boundary layer low-level clouds, were used to develop a parameterization scheme between visibility and a combined parameter as a function of both droplet number concentration Nd and liquid water content (LWC). The current NWP models usually use relationships between extinction coefficient and LWC. A newly developed parameterization scheme for visibility, Vis = f (LWC, Nd), is applied to the NOAA Nonhydrostatic Mesoscale Model. In this model, the microphysics of fog was adapted from the 1D Parameterized Fog (PAFOG) model and then was used in the lower 1.5 km of the atmosphere. Simulations for testing the new parameterization scheme are performed in a 50-km innermost-nested simulation domain using a horizontal grid spacing of 1 km centered on Zurich Unique Airport in Switzerland. The simulations over a 10-h time period showed that visibility differences between old and new parameterization schemes can be more than 50%. It is concluded that accurate visibility estimates require skillful LWC as well as Nd estimates from forecasts. Therefore, the current models can significantly over-/underestimate Vis (with more than 50% uncertainty) depending on environmental conditions. Inclusion of Nd as a prognostic (or parameterized) variable in parameterizations would significantly improve the operational forecast models.