In the summer of 1983 a group of scientists working in the fields of meteorology, biology, and remote sensing met to discuss methods for modeling and observing land-surface--atmosphere interactions on regional and global scales. They concluded, first, that the existing climate models contained poor representations of the processes controlling the exchanges of energy, water, heat, and carbon between the land surface and the atmosphere and, second, that satellite remote sensing had been underutilized as a means of specifying global fields of the governing biophysical parameters. Accordingly, a multiscale, multidisciplinary experiment, FIFE, was initiated to address these two issues. The objectives of FIFE were specified as follows: (1) Upscale integration of models: The experiment was designed to test the soil-plant-atmosphere models developed by biometeorologists for small-scale applications (millimeters to meters) and to develop methods to apply them at the larger scales (kilometers) appropriate to atmospheric models and satellite remote sensing. (2) Application of satellite remote sensing: Even if the first goal were achieved to yield a ''perfect'' model of vegetation-atmosphere exchanges, it would have very limited applications without a global observing system for initialization and validation.

Ascacresult, the experiment was tasked with exploring methods for using satellite data to quantify important biophysical states and rates for model input. The experiment was centered on a 15¿15 km grassland site near Manhattan, Kansas. This area became the focus for an extended monitoring program of satellite, meteorological, biophysical, and hydrological data acquisition from early 1987 through October 1989 and a series of 12- to 20-day intensive field campaigns (IFCs), four in 1987 and one in 1989. During the IFCs the fluxes of heat, moisture, carbon dioxide, and radiation were measured with surface and airborne equipment in coordination with measurements of surface and atmospheric parameters and satellite overpasses. The resulting data are held in a single integrated data base and continue to be analyzed by the participating scientists and others. The first two sections of this paper recount the history and scientific background leading up to FIFE; the third and and fourth sections review the experiment design, the scientific teams and equipment involved, and the actual execution of the experiment; the fifth section provides an overview of the contents of this special issue; the sixth section summarizes the management and resources of the project; and the last section lists the acknowledgments. ¿ American Geophysical Union 1992