Optical gravitational lensing experiment: OGLE-1999-BUL-19 – the first multipeak parallax event

We describe a highly unusual microlensing event, OGLE-1999-BUL-19. Unlike most standard microlensing events, this event exhibits multiple peaks in its light curve. The Einstein radius crossing time for this event is approximately 1 yr, which is unusually long. We show that the additional peaks in the light curve can be caused by the very small value for the relative transverse velocity of the lens projected on to the observer plane (). Since this value is significantly less than the speed of the orbit of the Earth around the Sun (v⊕∼ 30km s−1), the motion of the Earth induces these multiple peaks in the light curve. This value for is the lowest velocity so far published and we believe that this is the first multiple-peak parallax event ever observed. We also found that the event can be somewhat better fitted by a rotating binary-source model, although this is to be expected since every parallax microlensing event can be exactly reproduced by a suitable binary-source model. A face-on rotating binary-lens model was also identified, but this provides a significantly worse fit. We conclude that the most likely cause for this multipeak behaviour is parallax microlensing rather than microlensing by a binary source. However, this event may be exhibiting a slight binary-source signature in addition to these parallax-induced multiple peaks. With spectroscopic observations it is possible to test this ‘parallax plus binary-source’ hypothesis and (in the instance that the hypothesis turns out to be correct) to simultaneously fit both models and obtain a measurement of the lens mass. Furthermore, spectroscopic observations could also supply information regarding the lens properties, possibly providing another avenue for determining the lens mass. We also investigated the nature of the blending for this event, and found that the majority of the I-band blending is contributed by a source roughly aligned with the lensed source. This implies that most of the I-band blending is caused by light from the lens or a binary companion to the source. However, in the V band, there appears to be a second blended source 0.35 arcsec away from the lensed source. Hubble Space Telescope observations will be very useful for understanding the nature of the blends. We also suggest that a radial velocity survey of all parallax events will be very useful for further constraining the lensing kinematics and understanding the origins of these events and the excess of long events toward the bulge.