Analysis on Synoptic Climatology Characteristics and Forecast Methods of Fog in Hainan
|Keywords||Fog Climate Change trend Synoptic type Forecast method SVM Multiple linear regression|
Fog, as a disastrous weather and happens often in Hainan Island, fog gets more and more comprehensive attention. Fog influences communication and transportation, agricultural production and people’s good health more and more severe along with the development of Hainan economic. In order to enhance the forecast and early warning capability of fog weather, it is fundamental to strengthen researches on synoptic climatology characteristics and synoptic process mechanism of fog in Hainan.Different areas have different characteristics in Hainan Island just as that in china. Though there are only 34 thousand square kilometers in Hainan Island, the topography is compound with towering mountains in middle area and low plains around it. Because of the effect of topography and other factors, fog’s generation at different areas has its particularity in climate back ground, influencing synoptic systems, and so on. However, the researches on fog in Hainan are scarce till now, especially on technical studies of fog’s objective forecast.Based on the data including the original meteorological observational data from 1969 to 2008 in 18 counties of Hainan Island and the synoptic data as well as NCEP/NCAR reanalysis data, the climate characteristics and spatial and temporal distribution pattern of fog distribution at 18 stations in Hainan are studied by using EOF analysis. The results show that fog closely in associated with minimum temperature, relative humidity and other meteorological factors. The differences of atmospheric circulation conditions at 850hPa and 500hPa between the fog frequency years and rare years are also been compared. All kinds of fog occur under the stable atmospheric stratification condition, but the temperature inversion layer and the amount of water vapour vary from type to type. The radiation fog is influenced by diurnal temperature, especially in Haikou. In addition, a series of the operational forecasting methods and prediction system is established. Through this study, we conclude that:(1) There are most fog days in the central mountainous area, then in the northern coastland, and least in the southern coastland. Yearly fog days exhibited a decreasing trend. Analysis also indicates that yearly fog days was negatively closely correlated with average minimum temperature and positively correlated with relative humidity. Fogs formed more frequently and successively in autumn, winter and early spring. The regions with a higher altitude or a lower temperature had more fog days.(2) In the years when fog days was more than normal, the 850 hPa continental high was relatively weaker, and the 500 hPa subtropical high relatively weaker too, vice versa.(3)The statistics features and synoptic background of the heavy fogs in Hainan are distinguished. According to the synoptic systems, three types of fogs can be found in Hainan which are advection fog, radiation fog (including advection radiation fog and radiation advection fog) and front fog. The synoptic systems producing advection fog include seaside high ridge pattern and former front pattern; those producing radiation fog include continental high pattern, east high patter and even pressure pattern; that producing front fog is front pattern.(4) Cases study show fog forms when atmosphere stratification is stable, with inversion layer exists in common, though the inversion layer of the radiation fog is not obvious and the layer is higher for advection and radiation fog. There are weak convergence and ascending motion during fogs forming, with warm wet moisture transports to fog regions at same time. It is easy to satisfy for radiation fogs’forming because it is lighter and needs fewer moisture compared with the other two kinds of fogs. Font fogs need most moisture.(5)The factors are different to lead different kinds of fog and the main forecasting factors include total cloud cover, temperature, wind, humidity, synoptic pattern, and so on.(6) Finalizing optimal modeling according selected appraise standard by using SVM method after data choosing, forecasting factors confirming by filtering core indexes step by step. Then fog weathers in 24 hour are forecasted to Haikou and Wuzhishan with it. Multiple linear regression model is constructed and forecasting fog weathers in 24 hours to Haikou and Wuzhishan too. The results show that the two models have good forecasting technical which can be used in operational weather forecasting.The research fills the blank of studies about Hainan fog. The analysis of climate characteristics and synoptic type of fog in Hainan can provide statistical reference on fog weather’s analysis and forecasting. The researches on methods of fog forecast, coupling with practical operation, provide good reference on fog’s short forecast, and have good prospect of promotion and application.