Correlation Between GPS-derived Water Vapor And Regional PM2.5

Water vapor is an important component in the atmosphere.It is less abundant in the atmosphere but has a larger range of variation.Water vapor plays a leading role in various weather changes.Traditional atmospheric detection methods have great limitations in terms of continuity and accuracy,and GPS-derived water vapor technology has experienced all-weather,high precision,real-time and fast,free from weather conditions and low cost after undergoing many changes and developments in order to be widely used in various weather observation and forecasting work.With the rapid development of the economy in recent years,the emission of various pollutants has caused the environmental deterioration to intensify,and the air quality has been declining year by year.The haze is one of the important phenomena of air pollution,and its cause is that certain pollutants in the atmosphere are subject to certain conditions.The meteorological factors are coagulated or aggregated,and it is impossible to settle by self-settling in a short period of time to make the visibility worse.Atmospheric water vapor is one of the external meteorological factors that promote the formation of haze.This paper uses the data of GPS stations and national control weather station in five cities in China to establish a model of atmospheric precipitable water,atmospheric pollutant correlation and an air pollutant monitoring model based on water vapor inversion.The main contents include:(1)Acquisition of zenith precipitation PWV.Using GAMIT to process data from GPS stations in five cities in China to obtain PWV.Through the comparison of PWV acquired at the same time as the sounding station,the accuracy of PWV acquisition is compared and evaluated from the aspects of processing time,processing precision and processing complexity.(2)Correlation between PWV and PM2.5 of airborne particle in Xuzhou area.According to the regional overview of the selected research area and the extreme haze weather from month to month,the correlation and trend analysis of the atmospheric precipitation in Xuzhou and the PM2.5 data of the national control meteorological monitoring station were calculated by GAMIT.The comparison shows that the GPS retrievable atmospheric precipitation is positively correlated with PM2.5.The extreme weather reflected by the sample data is basically consistent with the historical weather information.(3)The establishment of retrieving water vapor air pollution monitoring model based on BP neural network.In order to obtain the model between the two moreaccurately,the neural network technology is used for learning prediction,and the PM2.5 monitoring model is established and compared with the air pollution particle data of the sounding station.The correlation coefficient is significantly higher than the linear regression model.The root mean square error is also significantly improved.The accuracy indexes indicate that the BP neural network optimization model is better than the linear regression model,and further validates the air pollution based on GPS water vapor inversion and BP neural network technology.Monitoring the feasibility of the model provides new possibilities for research in the field of atmospheric monitoring.