Application of the Fuzzy Sets Theory and FAO Method on Suitability and Clustering of Land Units in Marand region for Sunflower and Canola Products

In recent decades, increasing of population and limitation of arable lands have led to consider production level increasing and land suitability assessment. Several methods such as limitation (Maximum, Number and intensity), parametric (Story, square root) and a new fuzzy logic based methods have been developed to assess the suitability of lands. Classical methods of land suitability assessment are not able to express the continuous nature of soil properties and define the land suitability classes as non-collectible. Therefore, using of the fuzzy sets in studies of land suitability assessment can lead to a better understanding of them. Also fuzzy clustering and kohonen artificial neural network can be used for homogenous soils determination and requirements data completion. In this study, for comparison of the fuzzy set theory and FAO method, calculated correlation coefficients values between the land index and yield of the sunflower and canola by fuzzy method were (0.709 and 0.617, respectively), more than those with the Parametric method 0.709 and 0.617, respectively).These results demonstrate the high accuracy and potential of the fuzzy logic method in relation to the classical methods of land suitability assessment, and can be described by the capability of the fuzzy logic in expressing the continuous nature of the processes. Obtained results showed that representative profiles of 1 and 4 were homogenous based on landform clustering methods and these results could be used for information completion.