Predictability of an Advection Fog Event over North China. Part I: Sensitivity to Initial Condition Differences

Abstract

he predictability of a dense advection fog event on 21 February 2007 over north China (NC) is investigated with ensemble simulations using the Weather Research and Forecasting Model (WRF). Members with the best and worst simulation are selected from the ensemble, and their initial condition (IC) differences are explored. To test the sensitivity of fog simulation to those differences, the model is initialized with ICs that change linearly from the worst member to the best member, and the changes in simulated results are examined. The improvement in simulations due to the linear improvement of ICs is found to be monotonic. The IC differences at lower levels are of more influence to the simulation than IC differences at higher levels. By removing the IC differences of each meteorological variable individually, it is found that improvements in potential temperature and horizontal wind are more important than that of water vapor mixing ratio in this case. Additionally, the linear improvement in each meteorological variable also contributes monotonically to the simulated results. The budget analyses of the tendency of potential temperature and water vapor mixing ratio show that turbulence mixing and advection are the major factors contributing to the formation of fog. The correct initial temperature field ensures the formation and maintenance of an inversion, and the correct initial wind field ensures the correct transport of temperature and moisture in this case. Further discussion examines the reasons for the monotonic behavior in the simulation improvement.