Milliarcsecond angular resolution of reddened stellar sources in the vicinity of the Galactic center

Abstract

Aims. For the first time, the lunar occultation technique has been employed on a very large telescope in the near-IR with the aim of systematically achieving milliarcsecond resolution on stellar sources.

Methods. We have demonstrated the burst mode of the ISAAC instrument, using a fast read-out on a small area of the detector to record many tens of seconds of data at a time on fields of a few squared arcseconds. We used the opportunity to record a large number of LO events during a passage of the Moon close to the Galactic center in March 2006. We developed and employed for the first time a data pipeline for the treatment of LO data, including the automated estimation of the main data analysis parameters using a wavelet-based method and the preliminary fitting and plotting of all light curves.

Results. We recorded 51 LO events over about four hours. Of these, 30 resulted in sufficient quality to enable a detailed fitting. We detected two binaries with subarcsecond projected separation and three stars with a marginally resolved angular diameter of about 2 milliarcsec. Two more stars, which are cross-identified with SiO masers, were found to be resolved and in one case we could recover the brightness profile of the extended emission, which is consistent with an optically thin shell. The remaining unresolved stars were used to characterize the performance of the method.

Conclusions. The LO technique at a very large telescope is a powerful and efficient method for achieving angular resolution, sensitivity, and dynamic range that are among the best possible today with any technique. The selection of targets is naturally limited and LOs are fixed-time events; however, each observation requires only a few minutes, including overheads. As such, LOs are ideally suited to filling small gaps during the idle time between standard observations.