2002ApJ...572.1031A

We analyze PLANET and MACHO observations of MACHO 99-BLG-47, the first nearly normal microlensing event for which high signal-to-noise ratio data reveal a well-covered, short-duration anomaly. This anomaly occurs near the peak of the event. Short-duration anomalies near the peak of otherwise normal events are expected to arise both from extreme-separation (either very close or very wide), roughly equal-mass binary lenses and from planetary systems. We show that the lens of MACHO 99-BLG-47 is in fact an extreme-separation binary, not a planetary system, thus demonstrating for the first time that these two important classes of events can be distinguished in practice. However, we find that the wide-binary and close-binary lens solutions fit the data equally well and cannot be distinguished even at Δχ²=1. This degeneracy is qualitatively much more severe than the one identified for MACHO 98-SMC-1 because the present degeneracy spans two rather than one dimension in the magnification field and does not require significantly different blending fractions. In the Appendix, we explore this result and show that it is related to the symmetry in the lens equation.