An improved method of delineating rectangular management zones using a semivariogram-based technique

•We propose an improved method of rectangular management zone delineation.•The range of the semivariogram determines the size of the soil grid.•Optimal grid size well describes the soil homogeneity.•Optimal grid size reduces the time consumption of solving BILP.•Optimal grid size reduces the total variance of the obtained management zones.

Management zone delineation is of great importance for precision agriculture applications. Compared with oval management zones, rectangular management zones are more practical for variable rate technology and fertilization machinery, and they are also easy to use for farmers in developing regions. This research proposes an improved method for delineating rectangular management zones, which applies a semivariogram analysis to interpolating grid data with an optimal grid size. By using a well-designed grid size, optimal grids are considered when generating instances and solving binary integer linear programming (BILP) problems. These improvements greatly reduce the computational time as well as the total variance of the delineated management zones. The experimental results indicate that the proposed method provides a practical method of applying rectangular management zone delineation that performs better and is more efficient compared with conventional algorithms.