Stellar Spectral Classification And Parameter Measurement Based On Lick Indices On Hadoop

Celestial spectrum contains a wealth of astrophysical information. Through the analysis of spectra, people can get the physical information of celestial bodies, as well as their chemical composition and atmospheric parameters. With the implementation of LAMOST, SDSS telescope and other large-scale surveys, massive spectral data will be produced, especially along with the formal operation of LAMOST,2,000 to 4,000 spectral data will be generated each observation night. It requires more efficient processing technology to cope with such massive spectra. The subject is proposed under this background. Its goal is to study the automatic processing techniques for massive stellar spectra.Automatic processing of stellar spectra can be divided into two parts:one is the automatic classification of stellar spectra, and the other one is the automatic measurement of stellar atmosphere physical parameters. According to the relative intensity of the spectral lines and the continuous spectrum, as well as other features of the spectra, the stars are divided into O, B, A, F, Q K, M seven types. Distribution of Continuous spectrum and outline of spectral lines are determined by stellar atmosphere physical parameters including surface effective temperature (Teff), surface gravity (log g) and chemical abundances (Fe/H). Currently, most methods use wavelength and flux of spectra for spectral classification and parameter measurement. However, the dimension of the spectral data is high so that it needs a series of preprocessing such as normalization and reduction, which is a large amount of computation.In this paper, the methods of spectral classification and measurement of atmospheric parameters based on Lick indices have been studied. For massive spectra, the computation of Lick indices and the spectral classification using Bayesian decision method are implemented on Hadoop. With use of the high throughput and good fault tolerance of HDFS, combined with the advantages of MapReduce parallel programming model, the efficiency of analysis and processing for massive spectral data have been improved significantly.The main innovative contributions of this thesis are as follows.1. Using Lick indices as the characteristic, classify stellar spectra based on Bayesian decision and measure stellar atmospheric parameters based on kernel partial least squares regression method.2. Implementing parallel computation of Lick indices and parallel classification of stellar spectra using Bayesian based on Hadoop MapReduce distributed computing framework.