Perspectives on delineating management zones for variable rate irrigation Amir Haghverdi a,⇑ , Brian G. Leib b , Robert A. Washington-Allen c , Paul D. Ayers b , Michael J. Buschermohle b a Department of Biological Systems Engineering, University of Nebraska-Lincoln, Panhandle Research and Extension Center, 4502 Avenue I, Scottsbluff, NE 69361-4939, United States b Department of Biosystems Engineering & Soil Science, University of Tennessee, 2506 E.J. Chapman Drive, Knoxville, TN 37996-4531, United States c Department of Geography, University of Tennessee, Burchfiel Geography Building, Knoxville, TN 37996-0925, United States article info Article history: Received 27 April 2015 Received in revised form 20 June 2015 Accepted 23 June 2015 Keywords: Apparent electrical conductivity Integer linear programming Remote sensing Soil water retention Unsupervised clustering abstract This study aimed at investigating the performance of multiple irrigation zoning scenarios on a 73 ha irri- gated field located in west Tennessee along the Mississippi river. Different clustering methods, including k-means, ISODATA and Gaussian Mixture, were selected. In addition, a new zoning method, based on integer linear programming, was designed and evaluated for center pivot irrigation systems with limited speed control capability. The soil available water content was used as the main attribute for zoning while soil apparent electrical conductivity (ECa), space-borne satellite images and yield data were required as ancillary data. A good agreement was observed among delineated zones by different clustering methods. The new zoning method explained up to 40% of available water content variance underneath center pivot irrigation systems. The ECa achieved the highest Kappa coefficient (=0.79) among ancillary attributes, hence exhibited a considerable potential for irrigation zoning. Ó 2015 Elsevier B.V. All rights reserved. 1. Introduction 1.1. Precision farming and management zone delineation In conventional agriculture each field is considered as a uniform unit, by purposely ignoring the heterogeneity across the field, thereby decision-making is based on an estimation of average con- ditions. The motivation for site-specific farming was first addressed by researchers during the late 80s and early 90s (Arslan and Colvin, 2002). As such, precision agriculture (PA) methodology is a way to look at field management by taking the within field variation into account and incorporating that variability into management decisions. Within-field heterogeneity is caused by both temporal and spatial variation of a variety of factors such as climate, topography and biologic activity (Córdoba et al., 2013). A management zone (MZ) is a sub-region of a field that is rela- tively homogeneous with respect to soil-landscape attributes. It is expected that variable rate application across MZs will help by sav- ing the resources and optimizing yield (Schepers et al., 2004). Protecting the environment and keeping agriculture sustainable may also be achievable through precision farming. Sensor-based and map-based approaches are two major methods to practice variable-rate application. In the sensor-based method, a real time decision on application rate is made using data collected via sen- sors and a pre-developed application algorithms. In the map- based method, application maps are prepared using site-specific information such as yield data and soil data prior to implementa- tion. It is critical to select appropriate attribute(s) and method to delineate robust zones (Thöle et al., 2013). A field can be zoned based on a single soil-crop variable or mul- tiple attributes which are expected to affect yield (Khosla et al., 2010). Yield maps, topography, satellite photographs, canopy images and soil apparent electrical conductivity (ECa) are among suggested attributes to delineate MZs. Application of remote sens- ing is especially attractive because it is noninvasive and relatively inexpensive (Schepers et al., 2004). Yield maps are useful sources of information reflecting within-field variation. However, some dif- ficulties have been reported to delineate zones solely by yield maps (Khosla et al., 2010). Temporal inconsistency among yield maps from year to year is probably the main reason causing this problem. Schepers et al. (2004) reported that temporal climate variability in an irrigated cornfield significantly affects yield spatial variability from year to year. Combining yield data with other ancillary information or averaging yield data over years can help explain spatial variation better and in turn can provide more trus- table zones. Promising results have been reported by the studies that have utilized several years of yield data to create MZs. http://dx.doi.org/10.1016/j.compag.2015.06.019 0168-1699/Ó 2015 Elsevier B.V. All rights reserved. ⇑ Corresponding author. Tel.: +1 (308) 632 1246. E-mail address: ahaghverdi2@unl.edu (A. Haghverdi). Computers and Electronics in Agriculture 117 (2015) 154–167 Contents lists available at ScienceDirect Computers and Electronics in Agriculture journal homepage: www.elsevier.com/locate/compag