In astrophysics, MACHO is an acronym for MAssive Compact Halo Object, what is believed to be a component of dark matter. MACHO is also the name given to an observing project dedicated to the search for dark matter.

The nature of MACHOs themselves is unknown, though one view is that they are brown dwarf stars, or gas giant planets floating freely through the galaxy. These objects would be so faint that we couldn't detect them directly, but their collective mass may influence the kinematic behavior and rotation of the Milky Way and other galaxies.

Though we cannot see them shine by their own light, we can detect their presence by gravitational lensing. Imagine we point a telescope at a distant star, and measure that star's brightness over time. Now suppose that one of the MACHOs passes between us and that distant star. According to general relativity, the curvature of space around any mass can act like a lens for background light. Curved space "bends" the rays of light, sending them in a slightly different direction, much like glass lens does. If the MACHO is located close to the line of sight between us and the background star, we will see an increase in the light from that star as the MACHO bends the light around it. We see this effect change over time, because the background star, the MACHO, and the Earth all move through space over time. If the background star is not variable, we will first find it has constant brightness, followed by an increase in light, and then a decay down to the normal light level. These lensing events last from a few days to a few weeks, and do not repeat.

While there are lots of stars in the sky, and presumably a lot of dark matter as well, this happens very infrequently because space is so vast, and these objects are very far apart; the number of such microlensing(*) events is measured only in the hundreds since their first detection in the early 1990's. The way such events are detected is rather ingenious. You use a telescope (or two), and take an image of a dense star field a few times a night. For example, you can take pictures of the Milky Way's central bulge, or the Large and Small Magellanic Clouds. You then use a computer to measure the brightness of each star on the image. Over days and months and years you keep imaging the same patch of sky, and obtain a very long light curve of each star. You then use a computer to search each light curve for the characteristic shape of a microlensing event. It is a bit like searching for a needle in a haystack, but it works.

Several independent projects were undertaken to perform searches for dark matter in this way. Two of the most famous ones are the MACHO Project, a collaboration between many universities in the U.S., Canada, and Australia with a telescope in Australia, and the OGLE Project (Optical Gravitational Lens Experiment), a collaboration between the Carnegie Institution in Washington, Princeton University, and the Warsaw University Observatory in Poland with a telescope at Las Campanas Observatory in Chile.

While these projects were established to search for dark matter, one of the most important fringe benefits has been their detection of many thousands of variable stars in the Galactic Bulge, in globular clusters, and in the Magellanic Clouds. Since the light curves of millions of stars are measured every night, it is straightforward to search them for variability not due to gravitational lensing. Both the MACHO and OGLE projects have detected thousands of variable stars including RR Lyrae, Cepheid and delta Scuti variables, and binary stars. These stars have been used for many purposes, such studying the physics of stellar variability, and measuring the distance to the Milky Way's center and the Magellanic Clouds. Data on the Magellanic Clouds have also been used to make color-magnitude diagrams of millions of stars in the Magellanic clouds, to study its star formation history.

(*) We call these events microlensing, because we usually reserve gravitational lensing for what is done by massive clusters of galaxies on very large scales, or by single foreground galaxies on distant quasars.

Gritchka points out that MACHOs are much better dark matter candidates than WIMPs, for obvious reasons. :) Actually, a friend of mine thinks it's all bunk, and that topological defects are the way to go. I think cosmology gives me a headache....

My PhD dissertation: An Observational And Theoretical Study of High-Amplitude Delta Scuti Stars (June 2000)
I was an affiliate of the MACHO Project for part of my dissertation work.