Data Processing of the Middle and Upper Atmospheric Wind Retrieval Based on the Fabry‐Perot Interferometer

Abstract

Since the Doppler shifts of airglow orauroralemission lines due to neutral winds are extremely small, it is essential to determine the center position and radius of fringes with the uppermost accuracy. In this paper, a new method is presented for the center position and radius determination of incomplete interference fringes of Fabry-Perot Interferometer (FPI). Due to the data noise comes from FPI instrument and space environment, the data pre-processing was firstly carried out including mean filtering and wiener weightingadaptive filtering. Then, according to the Gauss distribution of fringe intensity, interference fringe peak coordinates were derived from both radius obtained using Gaussian fitting and fringe center coordinates. Subsequently, based on the peak coordinates, new center position and radius were estimated by using circular coordinate fitting with the least square method. Finally, the center coordinate and radius were determined by using the iterative operation between Gaussian fitting and circular coordinate fitting. The new method was applied to Fabry-Perot interferometer data processing for wind velocity estimation of the middle and upper atmosphere. The retrieval results are 96.9537 m/s (actual wind velocity ∼100 m/s) and 7.5282 m/s (actual wind velocity ∼10 m/s), lower by 2.977 m/s and 2.465 m/s, respectively. Compared with theoretical values, both of them are less than 3 m/s, which suggests a tentatively reliable and practicable method of circle center and radius determination for wind retrieval of the middle and upper atmosphere.