RELATING SPECTRAL OBSERVATIONS OF THE AGRICULTURAL LANDSCAPE TO CROP YIELD

Remote sensing and microscopy share several common concerns including wavelength and sensor selection, signal processing, and image analysis. For crop yield assessments, multispectral observations are acquired photographically, videographically, or with optical-mechanical scanners from aircraft and spacecraft. Sensors are chosen at wavelengths of high atmospheric transmission and maximum contrast between the soil background and the vegetation growing out of it. Vegetation indices have been developed that maximize the information about the photosynthetic size of the vegetation in the landscape and, hence, about crop stresses and yield. Three such indices that reduce the multispectral observations to a single numerical index are described and software for one general procedure that permits characterization of each major spectral component of multiband scenes is appended. Microscopists may encounter analogous situations for which the techniques developed in agricultural remote sensing can be useful.