Abstract
Surface reflectance and global irradiance variations across an imaged area determine the minimum signal difference from groups of pixels in different targets needed to discriminate them. The factors affecting such variations and their magnitudes are considered in this paper. The sensor output also depends upon the interaction of the spectral response of the sensor and the spectral upwelling radiance from the target. From a consideration of the spectral instrument responses of detectors in the Landsat multispectral scanners (MS), together with measured reflectance factors of a vegetative canopy stressed to different levels of severity, the feasibility of mapping and quantifying disease stress with Landsat vegetative indices (VINs) is demonstrated.