The Characteristics And Numerical Simulations Of Heavy Fong Along Nanjing To Shanghai Expressway

Based on the observed data of automatic meteorological monitoring system (AMMS) on Nanjing-Shanghai Expressway from 2006 to 2009,the climatic characteristics of fogs on the expressway was analyzed,the atmospheric circulation backgrounds during the fogs of different types occurred were reviewed and the relationship between the fogs and some meteorological factors were discussed. From the above study, many typical fog cases were sifted all fog events out. After the selection of different parameterization schemes and and the coupling of different phsical process schemes, WRF 2.2 (Weather Research and Forecast for Version 2.2) and the reanalyzed data (1°×1°) from NCEP (National Center for Environmental Predicting in USA) were utilized to simulate and diagnose these cases and some ideal simulation results were acquired.Considering the output simulations from WRF model,the different physical varible fields related to the fog cases were diagnosed and analyzed and the physical mechanisms on the engendering, developing, maintaining and disappearing of different fog processes were expounded. Some forecasting models of visibility based on the method of the supported vector machine (SVM) in different fog processes were constructed and the influence of different kernel functions and a Particle Swarm Optimization (PSO) algorithm on the forecasting effect were probed. The forewarning system of heavy fog weather on Nanjing-Shanghai Expressway was developed and it was used in a daily traffic meteorological business. The main research contents and conclusions in this dissertation were included as below:The observed data per minute came from the automatic meteorological monitoring system on Nanjing-Shanghai Expressway during 2006 to 2009 was processed and analyzed. Some climatic characteristics of fogs on the expressway was extracted as follows:(1)The fogs on the expressway had obvious local distribution that the foge events occurred more on the western part than the eastern part in general. (2) The frequency of heavy fogs in a year appeared in December, the next was January.The heavy fog events occurred in 6 months from October to March in the next year accounted for 76% of all fog heavy events and this period has been a obvious arising season of heavy fog. (3)In a typical heavy fog seadon, the frequency of heavy fogs emerging has been highest at 6:00 or 7:00 in the morning of a day and lowest after noon.A classification investigation on 172 heavy fog events which appeared on the expressway during June,2006 to May,2009 was carried out and the atmospheric circulation backgrounds of different fog types were analyzed. The analysis displayed as follows:(1) The heavy fog types occurred on the expressway inclued radiation fog,, local dumpling fog, mixed type from radiation fog and addvection fog, mixed type from radiation fog and dumpling fog, mixed type from addvection fog and dumpling fog and composited type from radiation fog, addvection fog and dumpling fog. (2) According to the synoptic situations of surface,850 hPa and 500 hPa, these heavy fog events were divided into four kinds of subjective types such as equal-pressure fog at the front of high-pressure area, equal-pressure fog at the front of low-pressure area, heavy fog at the back of high-pressure area which wet into the sea exactly, low-pressure or inverse trough fog at the front of high -pressure area. (3) In view of a method from the Lamb-Jenkinson objective classification that combined a objective circulation classification with some quantitative atmospheric circulation indexs, the atmospheric circulation backgrounds of these heavy fog events emerging were divided into 27 objective types.(4)When the heavy fog appeared on the expressway, the weather situaions of equal-pressure controlled the fog area in most cases, the pressure contours on sea level was relative sparse and the fair straight airflow prevailed in the northweatern direction.The analysis on the relationship between the occurrence of heavy fog eventsand some meteorological factors showed as follows:(1) Relative humidity, temperature and wind speed were the good indicative factors which could play an important role on the forecasting the fog events on the Shanghai-Nanjing Expressway. When the relative humidity at a monitoring station of AMMS was more than 90%, the air temperature ranged from 0℃to 15℃, and the wind speed was lower than 2m·s-1, a heavy fog would occur. (2)More than 78% of advection fogs had an precursor as elephant trunk-shaped with a sudden drop in the visibility evolving curve.WRF model for version 2.2 was used to simulate the atmospheric backgrounds of some sifted heavy fog cases. The simulations demonstrated as follows:(1)The content of air liquid water in the fog area,the vorticity field, the advective cooling and radiative cooling of the temperature field close to the ground during the nights were approached to the observed value. (2) Topographic condition has been one of important factors engendering local dumpling fogs and a digital elevation model (DEM) in ArcGIS 9.2, a software of geographical information system were used to analyze the change of surface slopes the southern part of Jiangsu Province during some dumpling fog events. It was found that the highway on the Zhengjiang Part situated in the environment of relative high terrain, obvious undulated landforms and apparent convergence of wind direction and speed. (3) The topographical and geomorphological sensitivity experiments appeared that the two factors were sensitive to the formation of dumpling fog. Undulated landforms has a promoting effect on the engendering and developing. (4) 7 groups of parameterization schemes were selected to simulate some heavy fog events and it was detected that the following parameterization schemes had ideal simulated outcomes on different physical variables and backgrounds fields in the heavy fog processes such as RRTM scheme reflected the balance of long-wave radiation, Dudhia scheme in the simulation of short-wave radiation, Kain-Frisch scheme of cumulus parameterization applied in the thick grid level, Monin-Obukhov scheme was optimum in the air layer near the ground, YSU scheme was proper to the boundary layer simulations, Lin scheme had the best effect on the fine physical processes and SLAB scheme was most accurate in the land suface physical processes.Some forecasting models of visibility based on the method of the supported vector machine (SVM) in different fog processes were constructed and the influence of different kernel functions. The researches result displayed as follows:(1) The forecasting models based on the SVM methods on the Shanghai-Nanjing Expressway had certain dependence on the selection of kernel functions. Cauchy kernel and radial kernel functions have the better forecasting capability in heavy fogs on the expressway according to the case experiments.The method was applicable in the forecasting on highways for its low error.(2)It was found that the Particle Swarm Optimization (PSO) algorithm could improve the computing efficiency better in generating parameters of SVM model than the exhaustive algorithm.