Multivariate spatial sample reduction of soil chemical attributes by means of application zones

Aim of study: In precision agriculture, the definition Application Zones (AZs) in agricultural areas consists delimiting area subareas with similar characteristics, using soil chemical attributes. To such end, use clustering methods is common. Therefore, AZs make up a database that can be used to target future sampling, thus seeking possible sample reduction. The objective this paper assess acquisition configurations, reduced size, contained application zones generated by spatial multivariate clustering. sampling protocol proposed work evaluated five (C-means, Fanny, K-means, Mcquitty, and Ward) for creation AZs, and, through these obtain configurations 50% 75% initial points.
 Area Commercial area, Cascavel, Brazil.
 Material methods: Data attributes from commercial were used, referring three soybean harvest years (2013-2014; 2014-2015; 2015-2016). considered dissimilarity matrix aggregates information about Euclidean distance between elements dependence structure attributes.
 Main results: results indicated division into two or aforementioned years, considering K-means method. Comparing all one, it was observed one proportionally 25% most effective configuration.
 Research highlights: showed reduce size.