Two-dimensional Spectrographic Data Pocessing And Software Development For LAMOST

The development over the last two decades of efficient and reliable multi-object fiber spectroscopy has resulted in enormous advances in modern astrophysics. The important astrophysical information is hidden in spectral lines of astronomical spectra, which is the important channel of obtain astrophysical information and is the important tool for research on astronomy and astrophysics. The key problem of the wide-field and large sample astronomy research is how to obtain vast astronomical spectra, which is related to astronomy and astrophysics forefront of many issues.The LAMOST(Large Sky Area Multi-Object Optical Fiber Telescope)is in the stage of development, and will yield very large amount of astronomical spectra. The question how to fast and effective cope with the anticipated flood of spectrographic data is raised. As an important component of the LAMOST data process system, the two-dimensional spectrographic data processing system provide a reliable source of data for further spectral recognition and analysis and for obtaining the ultimate physical information(e.g. redshift, radial velocity et al).According to the LAMOST observation process, the paper studied the formation Mechanism of the LAMOST two-dimensional spectrographic data, proposed the two-dimensional spectrographic data process method and developed the process software. The main work and innovations are listed as follows:LAMOST observation physical model and mathematical model have been built by the actual observation process. In order to research on LAMOST two-dimensional spectrographic data process method and verify the feasibility of the method, the simulated data has been produced according to the design parameters.Referring to SDSS, the two-dimensional spectrographic data process method of LAMOST was established. The version 1.0 and 2.0 of two-dimensional spectra data process software was also completed, which was tested with simulated data. The result indicated that the process could basically meet the designed demand of LAMOST.For the object spectra contaminated by the sky background, access to the true object spectra was affected by the sky subtraction approach which is an important and difficult problem of two-dimensional spectra data process. There are spatial variations in the sky background in observation area. But all current method ignored the spatial variations and combined the sampled sky spectra to replace the sky background. So the precision of sky subtraction was impacted, which is the upmost factor of observing depth. In the paper, the sky subtraction method was presented based on B-spline surface fitting sky spectra, and the simulation results shown that method can improve the precision of sky subtraction because of spatial variations in the sky background.A method was proposed to find and remove cosmic-rays from a single spectroscopic CCD images. First, a modified median filter is used to identify pixels which are likely to be contaminated by cosmic-rays, as noise candidates. Then an edge-preserving technique, call TV inpainting, was used to restore the noise corrupted pixels. Such a method is applied to spectroscopic CCD images. The result shows that the cosmic-ray can be well rejected by the method without obvious degradation on the pixel flux value.