| 其他摘要 | The sun is the nearest star to us; the change of the internal magnetic field leads to the eruption of the solar activities, and the activities of the sun is closely related to the living environment of the human being. Therefore, it is of great significance to monitor, predict and study its internal variation rules of the sun in real time. In addition to high spatial resolution solar imaging observations, high spectral resolution observations are also the important means of the solar observation.We can study the physical nature of the generation mechanism and evolution law of the solar atmospheric activity by the solar spectrum. The long time, high space and spectral resolution of two-dimensional imaging spectrum are the mainstream of the solar spectrum observation. The grating spectrometer is a kind of spectral imaging instrument to realize the two-dimensional spectral observation.The New Vacuum Solar Telescope (NVST) is the solar telescope with the largest aperture in China, equipped with many high-performance terminal equipment, such as the multi-channel High Resolution imaging system, multi-band and High-dispersion spectrometers and so on. At present, the imaging observation is at a mature stage, from observation to subsequent data processing, and the observation and data processing of spectrum is still in development stage. This paper is based on the spectral data of NVST observation, systematically studies the error source and error mechanism in the observation process, and establishes a complete high precision processing and methods for the NVST spectral data. The main research contents include the following five aspects：The first, the precise preprocessing of the solar spectra observed by the NVST. Through the analysis, it is found that there are many kinds of non-uniform responses that change with time in the NVST flat data, which leads to the traditional spectral flatting method introduce a large number of additional errors while processing the "false signal" of spectral data.Thus, we propose a method to separate the inhomogeneous factors in the flat field, and correct the factors which are changed with time, and then used them to the science data. In this way, the fixed inhomogeneity responses and some Time-varying factors can be corrected well, and no additional errors will be introduced into science data. After producing, we calculate the measuring accuracies of the Doppler velocities with the chromosphere and photosphere lines, separately; it can reach up to 100-200 m s-1 for Ha line center, and 100 m s-1 or photosphere lines. It is means that the characteristics of the chromosphere and the photosphere can be studied with reduced NVST spectral data co-spatially and co-temporally.The second, the processing of NVST two-dimensional scanning spectral data to improve the spatial resolution of 2-D composing monochromatic images. At present, the NVST 2D scanning spectral observation is carried with any aid of image stabilizer. There will be a large scale wobble motion is present during the spatial scanning, which will seriously decreases the spatial resolution of 2D composed monochromatic images. To solve this problem, we propose a method to calculate the displacements of solar image motion with the slit-jaw images of the TiO band, and the spectral data are corrected and rearrange by those displacements. After a detailed analysis and comparison, the corrected monochromatic images obviously have high spatial resolution.The third, the testing and deconvolution of the instrument profile of spectral data. The profile of the instrument will reduce the spatial and spectral resolution of the spectral data. We calculate the fixed part of the Point Spread Function (PSF) with the NVST spectral data, the direct derivation method and repeated trial-and-error method are used to obtain the average PSF of special direction, and the standard solar spectrum FTS matches the observed spectrum to get the average PSF in the dispersion direction. And then, the spectral data are deconvoluted in two directions. It can be seen from the results that the spatial resolution of 2D monochromatic images can be obviously improved by using 2D deconvolution in the special direction, and the result of the deconvolution in the dispersion direction shows that the adjacent and some weak spectral lines are distinguishable. It means the spectral resolution of the spectrum is raised.The fourth, the 2D monochromatic images are processed by the method of level1+. The average PSF reflects the average effect of the telescope system, while the influences of the atmospheric turbulence and the telescope are random. In order to completely eliminate them, a speckle reconstruction of the 2D monochromatic image is needed. Because the acquisition time of the slit-jaw image and the spectral data are not strictly synchronized in the scanning observations, we take the slit-jaw as an experiment to reduce by the methods of "De-stretching" and speckle reconstruction. Then the methods are applied to 2D monochromatic image of spectral composed, although there is obvious misplacement and time error in the image, the result of reconstruction also has some effect.The fifth, the denoising and compressing the NVST spectral data. By analyzing the characteristics of spectral data, we propose the method of principal component analysis (PCA) to reduce the noise and compression of NVST spectral data. The analysis of processing results shows that PCA technology can not only reduce the noise of the spectral data on the premise of preserving spectral information well, but also largely reduce the space of data storage, which is helpful for online publication in the future.With the methods and produces proposed in this paper, the NVST spectral data can be processed comprehensively and systematically, which is beneficial to fully exploit the observational performance of the NVST spectral instrument, and improve the utilization value of the spectral data. |
修改评论