Author
 |
Sadler, Edward |
|
RUSSELL, GRAHAM - UNIV OF EDINBURGH |
|
Submitted to: Site Specific Management for Agricultural Systems Proceedings
Publication Type: Proceedings Publication Acceptance Date: 1/29/1997 Publication Date: N/A Citation: N/A Interpretive Summary: Site-specific farming, or precision farming as it is coming to be named, has generated much interest in the farming industry. The innovation is that seed, fertilizer, or pesticide applications could be adjusted for a particular point in a field, while the equipment is moving. This is a great technological step past the uniform application to entire fields. Although it is expected (and sometimes hoped) that precision farming methods will help farmers both be more profitable and protect the environment, much research must be done. Because the study of small parcels of land (smaller than conventional research plots) means much more work, many plan to use models of crop growth to complement research. If a model were good enough, one could adjust the parameters of the model that describe the soil, and then make runs of the model for each spot in a field. The effect of recommending, say, a fertilizer application on each of the many spots that make up a field could be checked quickly and easily. However, the models that currently exist may not be suitable because they were built for other purposes, and may have limitations that prevent them from mimicking crop yield within a field. This review of the current literature examines the potential, the limitations, and the research needs for models to contribute to precision farming. Technical Abstract: Site-specific, or precision, farming is gaining acceptance by the full spectrum of farming interests. This growth is made possible by recent commercial availability of equipment, increasing awareness of inherent variability, recognition of consequences of uniform culture, and new techniques to acquire, interpret, and present the spatial data. One facet of site-specific management that has lagged, however, is the modeling of crop yield under spatially and temporally variable conditions. Optimally, modeling of crop growth and yield would be a useful tool for integrating spatially variable soil parameters, cultural practices, and fertilizer applications. However, most modeling so far has predicted mean yields rather than accounted for variability. Development of models has been aimed at estimates of large-scale effects of variable soils (soil family and series rather than the intra-map unit variation now needed). Attempts to make multiple runs with appropriate adjustments to soil parameters have met with mixed success. Quantitatively appropriate sensitivity to all important input parameters is required for full success, which knowledge has not been built into the models. No known crop models predict both the mean and the expected variability, though some hydrologic models do so. The paper summarizes existing work, addresses the requirements for modeling of crop yield in site-specific management, and proposes directions for future research. |
