Abstract
Experiments to determine grassland features using spectral characteristics were performed. Spectral reflectance was measured, using a hand-held spectrometer, at 102-151 random points in five Italian ryegrass(Lolium multifeorum Lam.)meadows under different vegetation conditions. The measurements were used for three applications:1)to visualize the relationship between red and near infrared spectral characteristics, 2)to determine the spatial pattern of aboveground biomass and apply it to gamma distribution, and 3)to determine the spatial pattern of plant coverage. The red-near infrared spectral reflectance relationship that was derived was a brief and clear index for the degree of dense or sparse coverage or the area of bare soil in grassland. Analysis of the biomass and coverage histograms determined different patterns for different plant growth or grassland conditions. Biomass distribution could be determined more accurately, both qualitatively and objectively, by applying the gamma distribution. Thses results suggest that several kinds of grassland features could be determined in a short period of time using spectral data measured with an hand-held spectrometer.