The Stellar Observations Network Group (SONG). A legacy archive of stellar time-domain spectroscopy

The Stellar Observations Network Group (SONG) network has operated for more than a decade, providing long-baseline high-cadence spectroscopic observations of bright stars and the Sun. The observations from 2014 through 2025 constitute a substantial archive of high-resolution spectra and precise radial velocities suitable for a broad range of time-domain stellar astrophysics. We present an overview of the current status, instrumentation, and scientific capabilities of the SONG network and describe the scope and accessibility of the SONG data archive (SODA). We further illustrate the breadth of science enabled by SONG observations, including asteroseismology, stellar variability studies, binary-star characterisation, and exoplanet research. We summarise the operational status and observing strategies of the SONG facilities, describe the available data products and archive infrastructure, and outline procedures for accessing archival observations and proposing new observations within the SONG community framework. The SODA archive currently contains more than 580,,000 spectra of 3091 stars obtained with SONG using either iodine-cell or thorium–argon wavelength calibration. The archive spans more than a decade of observations and includes extensive high-cadence time-series data for bright targets spanning a wide range of stellar types and variability classes. Access to the archive is available to members of the SONG community, which remains open to new participants who agree to follow the community policies. The SONG archive has developed into a major long-baseline resource for stellar spectroscopy and radial velocity time-series analysis. Continued expansion of the archive, together with coordinated observations obtained contemporaneously with TESS and future PLATO observations, is expected to enable new studies of stellar oscillations, variability, and exoplanet host stars through combined radial velocity and photometric analyses.