A Method of Sky-subtraction Based on Principal Component Analysis

Skylight is an important noise source in astronomical observations. The problem of sky-subtraction is an important factor that restricts the depth of multi-object fiber spectroscopic observations. The method of principal component analysis (PCA) comes from statistics, and it can be used to find the relations between the different skylight spectra and to obtain the skylight components contained in the object spectra. In order to study the sky-subtraction method of LAMOST, adopting a group of raw data from the Sloan Digital Survey System (SDSS), a simulation experiment is conducted, and the obtained result indicates that for the sky-subtraction, the PCA method is more effective than the SDSS reduction pipeline. In addition, a prospect is made for the application of the PCA method in LAMOST observations.