Global crop production forecasting data system analysis
The author has identified the following significant results. Findings led to the development of a theory of radiometric discrimination employing the mathematical framework of the theory of discrimination between scintillating radar targets. The theory indicated that the functions which drive accuracy of discrimination are the contrast ratio between targets, and the number of samples, or pixels, observed. Theoretical results led to three primary consequences, as regards the data system: (1) agricultural targets must be imaged at correctly chosen times, when the relative evolution of the crop's development is such as to maximize their contrast; (2) under these favorable conditions, the number of observed pixels can be significantly reduced with respect to wall-to-wall measurements; and (3) remotely sensed radiometric data must be suitably mixed with other auxiliary data, derived from external sources.