PHOTON STATISTICS LIMITS FOR EARTH-BASED PARALLAX MEASUREMENTS OF MACHO EVENTS

We analyze the limitations imposed by photon-counting statistics on ex tracting useful information about MACHOs from Earth-based parallax obs ervations of microlensing events. We find that if one or more large (s ay 2.5 m) telescopes are dedicated to observing a MACHO event for seve ral nights near maximum amplification, then it is possible, in princip le, to measure the velocity of the MACHO well enough to distinguish be tween disk and halo populations for events with omega A(m)(2) greater than or similar to 1 day(-1), where omega(-1) denotes the timescale of the event and A, denotes its maximum amplification. Thus, if it turns out to be possible to reduce all other sources of error to below that of photon statistics, it may be feasible to do useful Earth-based par allax observations for high amplification events.