Monitoring All Sky for Variability1

A few percent of all stars are variable, yet over 90% of variables brighterthan 12 magnitude have not been discovered yet. There is a need for an all skysearch and for the early detection of any unexpected events: optical flashesfrom gamma-ray bursts, novae, dwarf novae, supernovae, `killer asteroids'. Theongoing projects like ROTSE, ASAS, TASS, and others, using instruments withjust 4 inch aperture, have already discovered thousands of new variable stars,a flash from an explosion at a cosmological distance, and the first partialeclipse of a nearby star by its Jupiter like planet. About one millionvariables may be discovered with such small instruments, and many more withlarger telescopes. The critical elements are software and full automation ofthe hardware. A complete census of the brightest eclipsing binaries is neededto select objects for a robust empirical calibration of the the accuratedistance determination to the Magellanic Clouds, the first step towards theHubble constant. An archive to be generated by a large number of smallinstruments will be very valuable for data mining projects. The real timealerts will provide great targets of opportunity for the follow-up observationswith the largest telescopes.Comment: 6 pages, latex, minor changes, published in the Publications of the Astronomical Society of the Pacific as one of Millennium Essays: 2000, PASP, 112, 1281-128