The regression-basedmodel developed in this studywas empiricallybased and utilized official crop statistics from Kansas to derive a relationship between winterwheat specific remotely sensed parameters and reported yield statistics. One of the main drawbacks of such remotely sensed based empirical models for estimating crop yields has been that their application is valid only for the areas they have been calibrated for. This study developed a single generalized-model that was applied at the state level in Kansas andwas proven directly applicable to Ukraine.
The winter wheat regression-based model developed in this study assumes, like many other empirical remotely sensed based yield models (Ferencz et al., 2004) that the canopy vigor of winter wheat estimated by spectral NDVI measurements is directly related to final winter wheat yields. Specifically, NDVI measurements around the time of the maximum, which encompass the ‘critical period’ for grain production