1999ApJ...512..564D

We propose and evaluate the feasibility of a new strategy to search for planets via microlensing observations. This new strategy is designed to detect planets in ``wide'' orbits, i.e., with orbital separation, a, greater than ∼1.5R_E. Planets in wide orbits may provide the dominant channel for the discovery of planets via microlensing, particularly low-mass (e.g., Earth-mass) planets. This paper concentrates on events in which a single planet serves as a lens, leading to an isolated event of short duration. We point out that a distribution of events due to lensing by stars with wide-orbit planets is necessarily accompanied by a distribution of shorter duration events. The fraction of events in the latter distribution is proportional to the average value of q^1/2, where q is the ratio between planet and stellar masses. The position of the peak or peaks also provides a measure of the mass ratios typical of planetary systems. We study detection strategies that can optimize our ability to discover isolated short-duration events due to lensing by planets and find that monitoring employing sensitive photometry is particularly useful. If planetary systems similar to our own are common, even modest changes in detection strategy should lead to the discovery a few isolated events of short duration every year. We therefore also address the issue of the contamination due to stellar populations of any microlensing signal due to low-mass MACHOs. We describe how, even for isolated events of short duration, it will be possible to test the hypothesis that the lens was a planet instead of a low-mass MACHO, if the central star of the planetary system contributes a measurable fraction of the baseline flux.