A radial velocity survey for large Magellanic Cloud microlensed sources

We propose a radial velocity survey with the aim of resolving the current dispute on the LMC microlensing: in the pro-MACHO hypothesis, the lenses are halo white dwarfs or MACHOs in general; in the pro-star hypothesis, both the lenses and the sources are stars in various plausible structures of the Magellanic Clouds and the Galaxy. Star-star lensing should prefer sources at the backside of or behind the LMC disk, because lensing is most efficient if the source is located a few kpc behind a dense screen of stars, here the thin disk of the LMC. This signature of self-lensing can be searched for with a radial velocity survey, since the kinematics of the stars at the back can be markedly different from that of the majority of stars in the cold, rapidly rotating disk of the LMC. Detailed simulations of effect, together with optimal strategies for carrying out the proposed survey, are reported here. Assuming that the existing 30 or so alerted stars in the LMC are truly microlensed stars, their kinematics can test the two lensing scenarios; the confidence level varies with the still very uncertain structure of the LMC. Spectroscopy of the existing sample and future events requires about two or three good-seeing nights per year at a 4-8 m class southern telescope, either during the amplification phase or long after.