The spatial variability of biotic factors (e.g., weeds and pathogens) and abiotic factors (e.g., nutrients and water content) are likely to be mapped through remote sensing and therefore high spatial resolution satellite and airborne remotely sensed images can be used to delineate input prescription map for precision agriculture. First, the whole image need to be accurately geo-referenced and/or co-registered, ideally with a position error <0.3 to 0.5 m, or 1 pixel. Generally, once isolated the selected parcel image where site-specific operations are to be implemented, the following processes are required: a) assessing/ discriminating the agro-environment indicator in which to implement the variable rate input application; and b) splitting the parcel image into tiny rectangular plots (i.e. 25 to 200 m²); and c) designing the variable-rate input prescription map according with the selected decision taking criteria.

     Precision agriculture requires high spatial resolution images, pixel size <1.0-1.5 m. As spatial resolution increase the relative position error estimated in pixels increase. Normally, satellite and airborne images are commercially provided with position errors not acceptable for precision agriculture (e.g. >5 to 10 m). Therefore a geo-referenciation process to improve/ decrease the position error is needed. The aim of this paper is to discuss the position and the geo-referenciation and co-registration errors of high spatial resolution images for precision agriculture. Specific objectives are as following: 1) to indicate common position errors of commercially provided satellites; 2) to discuss the disadvantages of conventional image geo-registration using ground “hard-edge” points; 3) to resume the use of the semi-automatic geo-referenciation AUGEO® system for geo-registration; and 4) to indicate the prescription input map error as affected by the positioning error.