The Research On Wavelength Calibration For LAMOST

Guoshoujing Telescope is the Large Sky Area Multi-Object Fiber Spectroscopy Telescope (LAMOST). It is an innovative active reflecting Schmidt telescope in combining a large aperture with a wide field of view, and possesses the highest spectrum acquiring rate in the world. LAMOST, not only provides a world-class platform for astronomical observation, but also brings Chinese astronomy to a higher level in the fields of large scale spectra observation and large field of astronomy. Thus, LAMOST is an engineering of milestone meaning in the history of the development of Chinese astronomy.Astronomy researchers identify the celestial objects'physical structure,chemical composition and other information by analyzing their corresponding observed spectra. Spectrum analyzing needs to know the wavelength information of spectrum. Firstly, the light from celestial objects was observed by LAMOST. Then their corresponding two dimension fiber spectra were imaged on two CCD detectors of each spectrograph. Afterwards, their corresponding one dimension fiber spectra were acquired by flux extraction. For each one dimension fiber spectra, we only know its energy information about CCD pixels. The work described in this paper is to obtaining the relationship between spectral CCD pixels in the dispersion direction and their corresponding wavelengths. This is wavelength calibration in spectral data processing.This paper first introduces the construction background and system constitutes,observation model,data processing of LAMOST, then the works focus on the research of wavelength calibration of LAMOST. In summary, the main works are described as follows:1. Study the feature of one dimension arc spectra of LAMOST and gives two important algorithms in wavelength calibration: the gravity method to acquire the emission lines position and the polynomial fitting based on the least-squares to obtain the calibration coefficients. Then the paper details the wavelength calibration processing of LAMOST. Firstly, the initial wavelength calibration is implemented on a certain spectrum. Afterwards, the whole arc spectra calibration is completed based on the result of the initial wavelength calibration.2. As the disadvantage of the initial calibration coefficient optimization by enumeration is time consuming, we propose an effective optimization method for initial wavelength calibration of LAMOST based on PSO, to save optimization time and improve initial calibration quality. The method utilizes particle swarm to search the best initial wavelength calibration coefficients in the solution space near priori coefficients. Each particle generates its corresponding virtual arc spectrum by its position and the information of standard arc lines table. The cross correlation between the virtual spectrum and the actual observed one indicates the optimization quality of the current particle. Experiments demonstrate the effectiveness of the proposed method.3. We proposed a novel wavelength calibration method by combining arc and night sky lines for LAMOST, to improve calibration accuracy. The distributions of arc and sky lines are partly complementary in the wavelength range of LAMOST, thus those two kinds of lines could be employed jointly to set up the relationship between spectral CCD pixel positions and their corresponding wavelengths, the calibration result will more close to the real dispersion curve. Experiments demonstrate the effectiveness of the proposed method.