Delineation of Site-Specific Management Zones by Unsupervised Classification of Topographic Attributes and Soil Electrical Conductivity

Abstract

The objective of this research was to determine if unsupervised classification of topographic attributes and soil electrical conductivity could identify management zones for use in precision agriculture. Data collected in two fields located in central Missouri were used to test the proposed methodology. Principal component analysis was used to determine which layers of data were most important for representing within-field variability. Unsupervised clustering algorithms implemented in geographic information system (GIS) software were then used to divide the fields into potential management zones. Grain yield data obtained using a full-size combine equipped with a commercial yield sensing system and global positioning system (GPS) receiver were used to analyze the "goodness" of the potential management zones defined for each field. Principal component analysis of input variables for Field 1 indicated that elevation and bulk soil electrical conductivity (EC) were more important attributes than slope and Compound Topographic Index (CTI) for defining claypan soil management zones. The optimum number of zones to use when dividing a field may vary from year to year and was mainly a function of weather and the crop planted. The number of zones decreased if adequate moisture conditions were present throughout the cropping season (unpredictable) or if crops tolerant to water stress were planted (predictable). This classification procedure is fast, can be easily automated in commercially available GIS software, and has considerable advantages when compared to other methods for delineating within-field management zones.

Determination of potential management zones from soil electrical conductivity, yield and crop data

“…Several studies have indicated that homogenous management zones could be used as an alternative to grid soil sampling and to develop nutrient maps for variable rate fertilizer application (Khosla and Alley, 1999;Fleming et al, 2000a). Spatially coherent areas within fields may also be useful in relating yield to soil and topographic parameters for crop-modeling evaluation (Fraisse et al, 2001a While methods for delineating management zones vary widely in the information used, usually they are based on soil and yield information possibly over several years (Fraisse et al, 2001b;Fleming et al, 2000b). Many researches used the soil and/or relief information to define management zones.…”

“…Many researches used the soil and/or relief information to define management zones. For example, Fraisse et al(2001b) used a combination of topographic attributes and soil electrical conductivity (EC) to delineate management zones. Schepers et al (2004) aggregated the landscape attributes into management zones to characterize spatial variability in soil chemical properties and corn yields.…”

“…The most common is the use of cluster analysis. This can be used to identify areas that have similar landscape attributes, soil properties and plant parameters, to quantify patterns of variability and to reduce the empirical nature of defined management zones (Fraisse et al, 2001b). Stafford et al(1998) used fuzzy clustering of combined yield monitor data to divide a field into potential management zones.…”

Use of the farmer’s experience variable in the generation of management zones

“…The elevation attribute was selected to define the MZs for the three areas based on the results found by several authors, such as Bazzi et al (2015), Fraisse et al 2001, Jaynes et al (2005, Peralta and Costa (2013), Schenatto et al (2016), and Schepers et al (2004), which suggest that when the area is not flat, the variable elevation frequently has a spatial association with crop yield. …”

“…In the process of generating MZs, clustering methods have been widely used (FRAISSE et al, 2001;LI et al, 2007;REYNIERS et al, 2006;SCHENATTO et al, 2016;TAYLOR et al, 2003). The algorithm Fuzzy C-Means (BAZZI et al, 2013;LI et al, 2007;MORARI et al, 2009;MILNE et al, 2012;XIN-ZHONG et al, 2009) is a clustering method based on the fuzzy logic, defined by Zadeh (1965), which matches uncertainties associated with class and association boundaries (DOBERMANN et al, 2003).…”