Back to Volume
Paper: LAMOST DR5 Spectral Clustering for Stellar Templates Construction
Volume: 523, Astronomical Data Analysis Software and Systems XXVIII
Page: 91
Authors: Kong, X.; Luo, A.
Abstract: The large sky area multi-object fiber spectroscopic telescope (LAMOST) released the fifth spectral data (DR5) this year, containing 8,171,443 stars; 153,090 galaxies; 51,133 quasi-stellar objects (QSO); and 642,178 unknown types, all of which are classified by the LAMOST 1D pipeline. This pipeline is used for spectral analysis, aiming to determine the spectral type and redshifts of the spectra observed by LAMOST by matching them with spectral templates. Using template matching, we divide all the DR5 spectra into various groups according to their spectral type and signal-to-noise ratio. Then, we adopt k-means to build 500 cluster centers of spectra. After visual inspection, we select 197 centers that can be used as stellar templates for the pipeline. These templates are supposed to increase the number of types and the accuracy of the classification. In the end 19 cluster centers remain to be identified.
Back to Volume