Search

Yield is one of the primary concerns for any farmer since it is a key to economic prosperity. Yield productivity zones—that is to say, areas with the same yield level within fields over the long-term—are a form of derived (predicted) data from periodic remote sensing, in this study according to the Enhanced Vegetation Index (EVI). The delineation of yield productivity zones can (a) increase economic prosperity and (b) reduce the environmental burden by employing site-specific crop management practices which implement advanced geospatial technologies that respect soil heterogeneity. This paper presents yield productivity zone identification and computing based on Sentinel-2A/B and Landsat 8 multispectral satellite data and also quantifies the success rate of yield prediction in comparison to the measured yield data. Yield data on spring barley, winter wheat, corn, and oilseed rape were measured with a spatial resolution of up to several meters directly by a CASE IH harvester in the field. The yield data were available from three plots in three years on the Rostˇenice Farm in the Czech Republic, with an overall acreage of 176 hectares. The presented yield productivity zones concept was found to be credible for the prediction of yield, including its geospatial variations.

There is an increasing tension in agriculture between the requirements to assure full safety on the one hand and keep costs under control on the other hand, both with respect to (inter) national strategies. Farmers need to measure and understand the impact of huge amount and variety of data which drive overall quality and yield in their fields. Among others, those are local weather data, Global Navigation System of Systems data, orthophotos and satellite imagery, data on soil specifics etc. A strong need to secure Big Data arises due to various repositories and heterogeneous sources. Data storage and visualisation requirements are in some cases competing as they are a common interest as well as a threat that helps one part of a value chain to gain a higher profit. As demonstrated in this paper, handling (Big) data is therefore a sensitive topic, where trust of producers on data security is essential.