Numerical prediction of fog events over mountainous terrain: Sensitivity to model physics, data assimilation and visibility algorithms

The accurate prediction of fog events presents a significant challenge in numerical weather prediction. This problem becomes even notable in the regions of mountainous terrain, due to uncertainties in model physical parameterizations, initial conditions, and lower boundary conditions, as well as the intrinsic complexity in interactions between atmospheric boundary layer and surface of complex terrain.

In this presentation, challenges in the accurate prediction of fog events over mountainous terrain are demonstrated by evaluating the real-time forecasting results during the MATERHORN-Fog field program in January 2015 with the mesoscale community Weather Research and Forecasting (WRF) model and observations obtained during several intensive observing periods (IOPs).

Additional numerical experiments are performed to examine the sensitivity of numerical prediction of fog events to model physical parameterizations schemes (e.g., planetary boundary layer, radiation and microphysics), data assimilation and various visibility algorithms. Limitations and promises of fog prediction using the high-resolution numerical model with advanced data assimilation methods are discussed.