Detection of binary companions below the diffraction limit with lucky imaging

Abstract

Context. Binary stars are invaluable tools that can be used to precisely measure the fundamental properties of stars, to test stellar models, and further our understanding of stellar evolution. Stellar binarity may also play an important role in the formation and evolution of exoplanetary systems. Aims. We provide a technique for resolving intermediate-separation binaries stars with medium-sized telescopes (i.e. diameter less than or equal to 2.5 metres) at wavelengths around 825 nm in the super-resolution range (i.e. below the limit defined by the Rayleigh criterion). Methods. We combined two well-known algorithms that have been applied to reduce the halo in lucky imaging observations: COvariancE of Lucky Images and the Lucky Imaging Speckle Suppression Algorithm. We reviewed the fundamentals of both algorithms and describe a new technique called Lucky Imaging Super resolution Technique (LIST), which is optimized for peak highlighting within the first ring of the Airy pattern. To validate the technique, we carried out several observing campaigns of well-known binary stars with the FastCam instrument (FC) on the 1.52 m Carlos Sánchez Telescope (TCS) and 2.56 m Nordic Optical Telescope (NOT), both located at the Observatorios de Canarias (OCAN). Results. The projected angular separation between objects was resolved by applying LIST to FC data taken with TCS and NOT, with a result below 0.15''. It can go down to approximately 0.05'', given the limitations of the detector plate scale. This is, to our knowledge, the first time that binary companions with such small angular separations have been detected using only lucky imaging at optical wavelengths. The average accuracy achieved for the angular separation measurement is 16 ± 2 mas with NOT and is 20 ± 1 mas with TCS. The average accuracy obtained for the position angle measurement is 9.5º ± 0.3º for NOT and 11º ± 2º for TCS. We also made an attempt to measure the relative brightnesses of the binary components, obtaining results that are compatible with literature measurements. Using this comparison, the △m uncertainty obtained was 0.1 mag for NOT and 0.48 mag for TCS, although it should be noted that the measurements have been taken using slightly different filters. Conclusions. Lucky imaging, in combination with speckle suppression and a covariance analysis, can allow the resolution of multiple point sources below the diffraction limit of 2-m class telescopes. However, it should be noted that measurements in the super-resolution regime are less sensitive than those above the first Airy ring.