Mesoscale Numerical Study Of Dense Fog Event And Visibility Ensemble Forecasting: Case Study

A heavy fog event occurred in Anhui and Jiangsu province on 13-14 Dec.2007. In this study, a preliminary numerical simulation of this case was conducted by using the new generation Mesoscal model Weather Research and Forecast model (WRF), NCEP 1°×1°reanalysis data, routine observation data and the data from a comprehensive fog experiment carried out in Nanjing University of Information Science & Technology during the same period. The roles of dynamical factor, vapor factor and radiative cooling effect in the formation of a heavy fog were investigated by means of diagnostic analysis. The results show that the stable atmospheric stratification and abundant vapor supply led to the formation of fog. Long wave radiation cooling at the surface was of vital importance for the formation and the development of radiation fog. In fog formative stage the inversion was near surface and then got uplifted. Subsequent, the inversion layer was lifting, the existence of inversion at the top of fog layer played an important role in fog development and maintenance. The temperature profile in the fog layer changed slightly during the maintaining stage and after sunrise the inversion was destroyed since distinct solar radiation. Before fog formed, the lower layer was basically weak convergence region. There was downdraft under 950 hPa which was beneficial to the establishment and maintenance of stable stratification during fog episode. The water vapor budget for the fog was investigated based on simulated results of the WRF model. The research area was a rectangle contained Jiangsu and Anhui province, and the result showed that there was continuous water vapor import through the south and east boundary. During the fog there was convergence region of water vapor above 0.98a while divergence during the weakening period. The ground had a cooling effect on the surface layer because of receiving heat through radiation interchange, while it was reversed during the fog dissipation.An ensemble fog forecast system was designed based on WRF and a high resolution numerical 1-D model called PAFOG, initial perturbation conditions of PAFOG are obtained from the simulation of WRF. Comparison between prediction and observation was made during a typical radiation fog event that occurred in Nanjing during 13-14 December 2007. The results show the ensemble forecast system is better than single forecast. Using this method, the prediction for both visibility at surface and vertical structure of fog are better than those utilizing the model only initialized by radiosonde observation; the initial time of simulation played an important role on prediction, due to radiation fog mostly occurs at night, and the model will take some time to stabilize, the one-dimensional model started at 14:00 is the best choice, the standard deviation between ensemble member and observation is 0.417, and the absolute error of ensemble average is 0.228 km.