Method Research On Stellar Atmosphere Physical Parameter Extraction Based On Lick Indextransfer

At present, there are many large aperture with wide field of view survey telescopes in the world like the Sloan Digital Sky Survey (SDSS) Telescope of America, the Anglo-Australian Telescope of Australia and the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) of China, etc. They have produced massive spectral data. Researchers can not only determine atmospheric parameters and spatial distribution of stellar by observing the obtained stellar spectra, but also can combine age and kinematics information to obtain atmospheric parameters of different stellar population of the Milky Way, thus providing accurate constraints of the formation, structure and evolution model of the Milky Way. Researchers can also effectively trace the evolutionary history of the Milky Way since its formation, understand the nucleosynthesis theory, and examine the existing models of the universe by comparing stellar evolution models and observing results to promote researchers updating understanding of universe evolution. Stellar atmospheric parameters analysis is a fundamental approach to explore stellar, the Milky Way and even the evolution of universe. Such a massive spectral data proposed a higher requirement for processing quickly and effectively of spectral data.Stellar atmospheric physical parameters (effective temperature, surface gravity, and chemical abundances) are major factors leading to differences in stellar spectra. Automatic measurement of physical parameters of stellar atmosphere is an important research in large-scale survey telescope such as LAMOST automatic processing of spectral data.Lick index is a relatively wide spectral characteristics, each absorption line is named after the most prominent exponent. This index can ignore the error of correction and redshift traffic and have higher signal to noise ratio S/N, which makes Lick index have become an ideal method for measuring physical parameters of the atmosphere. This paper use Lick index, based on the characteristics of the mass spectra, respectively, by linear regression, artificial neural networks, support vector machine approach to design efficient algorithms to measure the physical parameters of the atmosphere. Firstly, we filter data of Kurucz spectrum, and then use them to train the three algorithms to find the best model. Finally, we test the model with the corresponding parameter of DR8spectrum. The results show that this method can effectively and accurately measured physical parameters of stellar atmosphere from low-resolution spectra and compared with current methods, this method can effectively estimate stellar atmospheric parameters faster and more accurate.Overall, the use of Lick line index for prediction of atmospheric physics parameters is feasible. For this paper, the three algorithms model needs to be further research to achieve a better fitting effect, which can be applied to the surveys project as LAMOST.