Incorporating a constrained optimization algorithm into remote sensing/precision agriculture methodology

For two decades now, the use of Remote Sensing/Precision Agriculture to improve farm yields while reducing the use of polluting chemicals and the limited water supply has been a major goal. With world population growing exponentially, arable land being consumed by urbanization, and an unfavorable farm economy, farm efficiency must increase to meet future food requirements and to make farming a sustainable, profitable occupation. “Precision Agriculture” refers to a farming methodology that applies nutrients and moisture only where and when they are needed in the field. The real goal is to increase farm profitability by identifying the additional treatments of chemicals and water that increase revenues more than they increase costs and do not exceed pollution standards (constrained optimization). Even though the economic and environmental benefits appear to be great, Remote Sensing/Precision Agriculture has not grown as rapidly as early advocates envisioned. Technology for a successful Remote Sensing/Precision Agriculture system is now in place, but other needed factors have been missing. Commercial satellite systems can now image the Earth (multi-spectrally) with a resolution as fine as 2.5 in. Precision variable dispensing systems using GPS are now available and affordable. Crop models that predict yield as a function of soil, chemical, and irrigation parameter levels have been developed. Personal computers and internet access are now in place in most farm homes and can provide a mechanism for periodically disseminating advice on what quantities of water and chemicals are needed in specific regions of each field. Several processes have been selected that fuse the disparate sources of information on the current and historic states of the crop and soil, and the remaining source levels available, with the critical decisions that farmers are required to make. These are done in a way that is easy for the farmer to understand and profitable to implement. A “Constrained Optimization Algorithm” to further improve these processes will be presented. The objective function of the model will be used to maximize the farmer's profit via increasing yields while decreasing environmental damage and decreasing applications of costly treatments. This model will incorporate information from Remote Sensing, from in-situ weather sources, from soil history, and from tacit farmer knowledge of the relative productivity of selected “Management Zones” of the farm, to provide incremental advice throughout the growing season on the optimum usage of water and chemical treatments.