2013A&A...555A..16T

OGLE-2002-BLG-360 was discovered as a microlensing candidate by the OGLE-III project. The subsequent light curve, however, clearly showed that the brightening of the object could not have resulted from the gravitational microlensing phenomenon. We aim to explain the nature of OGLE-2002-BLG-360 and its eruption observed in 2002-2006. The observational data primarily come from the archives of the OGLE project, which monitored the object in 2001-2009. The archives of the MACHO and MOA projects also provided us with additional data obtained in 1995-99 and 2000-2005, respectively. These data allowed us to analyse the light curve of the object during its eruption, as well as the potential variability of its progenitor. In the archives of several infrared surveys, namely 2MASS, MSX, Spitzer, AKARI, WISE, and VVV, we found measurements of the object, which allowed us to study the spectral energy distribution (SED) of the object. We constructed a simple model of a star surrounded by a dusty envelope, which was used to interpret the observed SED. Our analysis of the data clearly shows that OGLE-2002-BLG-360 was most probably a red transient, i.e. an object similar in nature to V838 Mon, whose eruption was observed in 2002. The SED in all phases, i.e. progenitor, eruption, and remnant, was dominated by infrared emission, which we interpret as evidence of dust formation in an intense mass outflow. Since 2009 the object has been completely embedded in dust. We suggest that the progenitor of OGLE-2002-BLG-360 was a binary, which had entered the common-envelope phase a long time (at least decades) before the observed eruption, and that the eruption resulted from the final merger of the binary components. We point out similarities between OGLE-2002-BLG-360 and CK Vul, whose eruption was observed in 1670-72, and this strengthens the hypothesis that CK Vul was also a red transient.