It is critical that American producers have new technologies that optimize their input costs with respect to profits. "Precision agriculture" deals with optimizing inputs on each small unit of an agricultural field as follows: (1) collection of detailed and site-specific data, (2) knowledge of relationships in the data between inputs and profits, and (3) ability to apply inputs on a spatially variable basis. The required steps in developing precision agriculture commercially are to demonstrate the incorporation of a number of precision-agriculture technologies onto a commercial field, and to conduct intensive and highly organized data collection on other fields as a validation of the technologies. Such an effort would also provide a comprehensive and extremely high-quality database from which the technologies could be validated for near-term commercialization, and it would increase the "comfort level" among clientele.

1. To evaluate remote sensing on cotton and soybeans for detecting variations in pest (weed, insect, disease) populations, soil and water profiles, and crop vigor, development, and maturity.
2. To refine procedures for and promote adoption of precision crop management based on remote-sensing/spatial technologies on a commercial farm operation.
3. To develop a comprehensive database, via an interdisciplinary team approach, that can be used to validate applications of remote sensing in agricultural systems.

"Truthing teams" will "ground truth" remote-sensing data within the designated areas (Stoneville and Starkville/Brooksville, and possibly one or two other sites). Ground truthing will involve intensive collection and georeferencing of agronomic and biological data. Zones delineated with remote-sensing data will be used to develop sampling profiles of selected cotton and soybean fields within the study areas. Representative samples of these zones will be made, with special consideration given to anomalies in the data. Hyperspectral data from a hand-held spectrophotometer will also be collected at each sample site. Ground-based data will be related to remote-sensing data so that remote sensing.

As regards the commercialization effort of this project, remote-sensing based precision-agriculture technology will be practically implemented. Specific plans are 1) to use ground-based weed-maps to enhance remote-sensing capabilities in weed detection, 2) to develop sampling profiles that efficiently define arthropod distributions in relation to spatial variability apparent in remote-sensing data, 3) to improve and integrate the methods by which remote-sensing based prescriptions of various management resources are applied, and 4) to utilize spatial technologies in real-world production fields for evaluation, validation, and demonstration of their value to end users. To this end, remote-sensing data will be used to prescribe application rates for insecticides related to the control of certain pests. A commonly used insecticide application system will be modified such that it has the capacity to respond very quickly to computer commands for a change in application rate. The insecticide will be applied, and analyses of overall system accuracy, efficiency, and cost savings will be conducted. Spatially variable applications of insecticides will be made on several production fields based on multispectral remote-sensing data from a commercial source. These efforts will enlist the collaboration of research, extension, industry, grower and crop consulting personnel.