Assessment of Field Spatial and Temporal Variabilities to Delineate Site-Specific Management Zones for Variable-Rate Irrigation

Abstract

Quantification and analysis of field variability are important initial steps in delineating potential variable-rate irrigation (VRI) management zones within an agricultural field. This study seeks to utilize variability in soil physical and chemical properties and in field elevation across a 27-ha field at the Alberta Irrigation Technology Centre (AITC) in southern Alberta, Canada, to define site-specific management zones. All geospatial data were collected during the 2013 and 2014 growing seasons. A stepwise multivariate regression approach was used to investigate how multiple measured parameters affect wheat yield. An unsupervised clustering algorithm, fuzzy c-means, was used to delineate the irrigation management zones. Fuzziness performance index (FPI) and normalized classification entropy (NCE) were used as verification criteria to determine the optimal number of management zones. Results revealed that soil electrical conductivity (EC) and field elevation were better suited for management zone delineation. Three management zones were identified based on the verification criteria using EC and field elevation variables. Measured crop yield differences corresponding to the three noncontiguous management zones were significant. The study area was categorized as low, medium, and high productive zones. The maximum wheat yield (4.80 t ha−1) was attained in the high-productivity zone; the lowest (2.22 t ha−1), in the low-productivity zone.