A method for predicting short‐time changes in fine particulate matter (PM2.5) mass concentration based on the global navigation satellite system zenith tropospheric delay. (26th December 2019)

• Record Type: Journal Article
• Authors: Guo, Min; Zhang, Hanwei; Xia, Pengfei
• Is Part Of: Meteorological applications. Volume 27:Number 1(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2019-12-26
• Subjects: fine particulate matter (PM2.5) -- orthonormal wavelet db5 -- regression analysis -- zenith tropospheric delay (ZTD); Meteorology -- Periodicals; Meteorological services -- Periodicals; 551.5
• DOI: 10.1002/met.1866
• Languages: English
• ISSNs: 1350-4827

Abstract

Abstract: In this study, a method of haze prediction has been developed based on zenith tropospheric delay (ZTD). The relationship between ZTD and fine particulate matter (PM2.5, the main component of haze) during hazy periods in Beijing from 2015 to 2018 was analysed. The correlation between ZTD and the PM2.5 series was analysed based on data from three hazy periods with relatively stable weather conditions, no heavy rainfall and relatively continuous data, and the correlation coefficient between ZTD and the PM2.5 series was maintained in the range [0.5207, 0.7883]. The analysis showed a strong correlation between them based on data from three hazy periods with relatively stable weather conditions. To enhance the correlation between ZTD and the PM2.5 sequence, the standard orthogonal wavelet Daubechies (db5) method was used in this research. By removing the influence of the high frequency signal using the method, the ZTD and PM2.5 sequences were reconstructed using the fourth‐layer low frequency coefficients, their correlation coefficient was maintained in the range [0.5540, 0.9067] and the percentage range of the correlation coefficient was increased to [2.46, 16.69]. Finally, the reconstructed PM2.5, ZTD, relative humidity, average wind speed and NO2 series through db5 were used to establish a multiple regression model to predict the change in the PM2.5 mass concentration. The experimental results show that the multivariate regression model after wavelet analysis is superior to the traditional multivariate regression method in predicting the short‐term change of PM2.5 through four statistics values ( R 2, F, P, S 2 ) of the regression model and that it is effective and feasible for predicting short‐term haze. Abstract : The relationship between fine particulate matter (PM2.5 ) and air quality index (AQI) in Beijing for October 1, 2015 to February 29, 2016; October 1, 2016 to February 28, 2017; and October 1, 2017 to February 28, 2018.