Identification of quasars that are variable over long timescales from infrared surveys

Context. Quasars are variable, and their variability can both constrain their physical properties and help to identify them. Aims. We aim to look for ways to efficiently identify quasars exhibiting consistent variability over multi-year timescales based on a small number of epochs. Using infrared (IR) is desirable to avoid bias against reddened objects. Methods. We compared the apparent brightness of known quasars that have been observed with two IR surveys, covering up to a twenty-year baseline: the Two Micron All Sky Survey (2MASS; 1997-2001) and the VISTA Hemisphere Survey (VHS; 2009-2017). We looked at the previous studies of the selected variable quasars to see if their variable behaviour is known, and when available we used multi-epoch monitoring with the Zwicky Transient Facility (ZTF) to obtain a measure of optical variability of individual objects. Results. We built a sample of nearly 2500 quasars that show statistically significant variability between the 2MASS and VHS. About 1500 of these come from the new Quaia sample based on Gaia spectra, and about 1/3 of these have hardly been studied. The Quaia sample constitutes the main product of this work. Based on ensemble variability and structure function analysis, we demonstrate that the selected objects in our sample are representative of the typical quasar population and show behaviour consistent with other quasar samples. Our analysis strengthens previous results showing, e.g., that variability decreases with the rest-frame wavelength and that it exhibits peaks for certain absolute magnitudes of the quasars. Similarly, the structure function shows an increase in variability for rest-frame time lags below ∼1500 d and a decrease for longer lags, just as in previous studies. Conclusions. Our selection, even though it is based on two epochs only, seems to be surprisingly robust, showing up to ∼11% contamination by quasars that show stable non-variable behaviour in the ZTF.