Research And Application Of Meteorological Correlation Network Model Based On Complex Network

The meteorological system is composed of many meteorological elements.There are complex interactions between meteorological elements.At the same time,one of the main reasons for changing of air quality is meteorological factors,that is,the changing and interactions of various meteorological elements.Therefore,it is very necessary to study the correlation between meteorological elements and the correlation between air quality and meteorological elements.But the changes between meteorological elements are complex and diverse.Complex network,as a theoretical method that can describe the relationship between multiple elements,can explain the inherent regularity of the fluctuation between meteorological elements,and highlight its essential characteristics.Previous research on Meteorological network lacks of correlation of meteorological elements and neglects the influence of internal correlation to air quality.For example,only considering the fluctuation of temperature as a single variable in meteorological elements,neglecting the correlation of correlation with other meteorological elements(such as temperature and pressure),nor the influence of the variation of its own correlation on the air quality.In order to solve above problems,firstly we use coarse-grained method to process data.Secondly,we constructed the correlation network of meteorological elements based on data mining algorithm and complex network theories,and then studying the effect of geographical and seasonal factors on the correlation network.Finally,the correlation network of AQI and meteorological elements was constructed for discussing major factors affecting AQI in different seasons.The mainly work of this paper has been summarized as follows:A meteorological element correlation network model was established.At the same time,the basic topology properties of the network and the practical significance of each property represented in the network were studied.The temperature and pressure correlation network of 22 cities in China is studied and analyzed.The temperature pressure correlation network of four seasons in spring,summer,autumn and winter is set up in each city,and the topological properties of temperature pressure network in different cities are analyzed by season.The 22 cities are divided into 6 regions to study the effects of different geographical locations on the topological properties of temperature and pressure networks.The correlation network of AQI and meteorological elements was constructed.Inthis network,only symbolic sequences in the same frequent itemsets are connected to each other.We analyze the meteorological fluctuation that connect with the air quality fluctuation,and find out the meteorological fluctuations affecting air quality in different seasons.Taking Shanghai city as an example,we search frequent itemsets from symbol sequence first and then construct the correlation network of AQI and meteorological elements based on frequent itemsets.Finally we found that relative humidity fluctuations,temperature fluctuations,and pressure fluctuations accompanied fluctuations in AQI during the four seasons.