Helicopter-based radiometric measurements of grassland sites were acquired during the First International Satellite Land Surface Climatology Project (ISLSCP) Field Experiment (FIFE) to quantify the spatial and spectral variations in surface reflectance contributing to variations in albedo and for comparison within surface fluxes. The helicopter instrumentation consisted of an eight-channel modular multiband radiometer (MMR), video, and photographic cameras. Seasonal and spatial variability of the entire FIFE study area were characterized using 20 MMR data sets from five FIFE sites during the four best (''Golden'') data collection days of 1987. The means and coefficient of variations for each spectral band were calculated and analyzed as a function of spatial and seasonal domains. Linear models describing simple ratio (SR) and normalized difference vegetation index (NDVI) as functions of several surface variables were tested. All eight MMR band contributed information beneficial for characterizing spatial and temporal variations, and the most useful bands were the red (MMR3, 0.63--0.68 μm) and the third middle infrared (MMR7, 2.08--2.37 μm).

The blue band (MMR1, 0.45--0.52 μm) proved to be valuable, whereas the often utilized near infrared band (MMR4, 0.75--0.88 μm) was among the least important bands for discriminating landscape variations. The SR was more sensitive than the NDVI to variations in the Konza grassland at all scales: intrasite and intersite, between managment practices of burning and grazing, and over time. Furthermore, the relationship of the NDVI with leaf area index was curvilinear, where the relationship of the SR was linear at low-leaf area index and was improved by the addition of a second variable, the percent of green dry weight biomass. Improvements in data collection methodologies are recommended for furture large scale experiments such as the Boreal Ecosystem and Atmosphere Study (BOREAS). ¿American Geophysical Union 1992