Parametric-based neural networks and TOPSIS modeling in land suitability evaluation for alfalfa production using GIS

Land evaluation is the process of predicting land use potential on the basis of its attributes. In the present study, the qualitative land suitability evaluation using parametric based neural networks and TOPSIS models was investigated for irrigated alfalfa production in Joveyn plain, Northeast of Iran. Some twenty-six land units were studied at the study area by a precise soil survey and their morphological and physicochemical properties. Our results indicated that the most limiting factor for alfalfa cultivation in the study area was soil fertility properties. The values of land indexes by neural networks model ranged from 46.39 in some parts in east and west to 75.91 in the middle parts of the study area, which categorized the plain from moderate (S3) to high (51) suitable classes. The TOPSIS preference values for alfalfa cultivation in the study area varied between 0.388 and 0.773 which classified from moderate to very high classes. The coefficient of determination revealed a high correlation between the output results of two models (R-2 = 0.961).