Recently, Dr. Shuang Gao (a LAMOST Fellow), Dr. Chao Liu (Associate professor) and their team found that incidence of binary cases in FGK stars is a function of stellar effective temperature and metallicity by using the SDSS and LAMOST spectral survey data.

Binary stars are widely distributed throughout the Milky Way. The frequency of binary stars is a key factor in astronomy for creating Galactic models and for studing stellar populations in the Milky Way.

They estimated the frequency of binaries from 5700 stars observed by SDSS and 5200 stars observed by the LAMOST survey. The radial velocities of stars are not constant, because the difference between two successive measurements of radial velocity is due to two components: one part is error in the radial velocity (RV) (which follows a Gaussian distribution) and the other one is orbital motion of the binary system (which does not follow a Gaussian distribution). The region outside of the Gaussian distribution is dominated by binary stars, and this region needs to be determined by our method.

These researchers built models of both RV errors and orbital motions of binaries to simultaneously fit components of single and binary stars based on variations in RV from SDSS and LAMOST data. Because observed samples from SDSS and LAMOST are distributed unequally in the Milky Way, they had to correct the bias of the sample in order to apply the results of the sample to our Galaxy. The fractions of binaries from SDSS and LAMOST in the Milky Way are obtained by correcting the bias of the sample being studied. The results are 43% from SDSS and 30% from LAMOST. Different coverage of the sample caused this difference in rates because there is a higher fraction of metal-rich stars in the LAMOST sample. To study the effects of stellar types, the frequency of binaries is re-estimated in intervals of temperature and metallicity. They found that frequency decreases with reduced temperature (left panel of the figure), and metal-poor stars have more binaries (right panel of the figure). These researchers discussed possible interpretations as follows: (1) The same binary fraction initially exists, but the measurable period of old (metal-poor) binaries has a limit. They could only can detect binaries with periods shorter than 1000 days due to the accuracy of RV measurements. A binary with a shorter period is easier to detect by a spectral survey. (2) More companions are likely to form in an environment with a high gas-dust ratio around metal-poor stars.