Integration of ANP and Fuzzy set techniques for land suitability assessment based on remote sensing and GIS for irrigated maize cultivation

Land suitability assessment can inform decisions on land uses suitable for maximizing crop yield while making best use, but not impairing the ability of natural resources such as soil to support growth. We assessed the suitability of maize to be produce in 12,000 ha land of Dasht-e-Moghan region of Ardabil province, northwest of Iran. Suitability criteria included soil depth, gypsum (%), CaCO3 (%), pH, electrical conductivity (EC), exchangeable sodium percentage (ESP), slope (%) and climate data. We modified and developed a novel set of techniques to assess suitability: fuzzy set theory, analytic network process (ANP), remote sensing and GIS. A map of suitability was compared a map created using a traditional suitability technique, the square root method. The coefficient of determination between the land suitability index and observed maize yield for square root and ANP-fuzzy methods was 0.747 and 0.919, respectively. Owing to greater flexibility to represent different data sources and derive weightings for meaningful land suitability classes, the ANP-fuzzy method was a superior method to represent land suitability classes than the square root method.