2014MNRAS.437.1318A

While Kepler has pushed the science of asteroseismology to limits unimaginable a decade ago, the need for asteroseismic studies of individual objects remains. This is primarily due to the limitations of single-colour intensity variations, which are much less sensitive to certain asteroseismic signals. The best way to obtain the necessary data is via very high resolution ground-based spectrography. Such observations measure the perceived radial-velocity shifts that arise due to stellar oscillations, which exhibit a much better signal-to-noise ratio than those for intensity observations. Stellar Observations Network Group (SONG), a proposed network of 1m telescopes with spectrographs that can reach R = 110 000, was designed with this need in mind. With one node under commissioning in Tenerife and another under construction in China, an analysis of the scientific benefits of constructing additional nodes for the network is warranted. By convolving models of asteroseismic observables (mean densities, small separations) with the anticipated window functions for different node configurations, we explore the impact of the number of nodes in the SONG network on the anticipated results, across the areas of the Hertzsprung-Russell diagram where solar-like oscillations are found. We find that although time series from two SONG nodes, or in some cases even one node, will allow us to detect oscillations; the full SONG network, providing full temporal coverage, is needed for obtaining the science goals of SONG, including analysis of modes of spherical harmonic degree l = 3.