应用气象学报 (Jan 2020)
Refractive Index Quality Control and Comparative Analysis of Multi-source Occultation Based on Sounding Observation
Abstract
Three occultation refractive index data of COSMIC, Metop-A and FY-3C from 1 September 2017 to 31 August 2018 are divided into four climate zones according to climate characteristics, so that the dataset is more consistent with the normal distribution. Occultation refractive index data and the deviation from the sounding are statistically analyzed. According to statistical results, the occultation refractive index data are quality controlled. Results show that the double-weighted mean and the double-weight standard deviation of three occultations refractive index are relatively close, and the overall trends are gradually decreasing with height. There are differences between the double-weighted mean and the double-weighted standard deviation in four climate zones. The subtropical monsoon climate zone is rich in water vapor in the lower troposphere, so its double weight mean and standard deviation are larger than those in other three climate zones. In the statistical calculation of the deviation from the sounding, the COSMIC is negative below 5 km, the subtropical monsoon climate zone has large deviation. The deviation of FY-3C in subtropical monsoon climate zone is positive below 2 km, negative between 2-6 km, positive again above 6 km, and the overall deviation is within 2 N. According to different statistical characteristics of deviations in four climate zones, different quality control standards are formulated, and error data and suspicious data are screened out. The threshold value of the correlation coefficient between the occultation and the sounding refractive index determined by statistical calculation is 0.44, and suspiciousness smaller than 0.44 is taken as wrong data. Below 1 km, the proportion of error data in quality control of occultation itself is 5%-10%. Above 1 km, within 5%, three occultations are relatively close. After importing the sounding observation reference, quality control data of Metop-A occultation in the plateau mountain climate area are increasing, while the others are about 6%. Comparing the correlation coefficient between occultations and the sounding before and after quality control, the value is smaller before the quality control. After error data being eliminated, the correlation between the occultations and the sounding refractive index is improved, mostly above 0.9. The control is effective, and the quality of the occultation data is improved.
Keywords