Abstract
The main viewpoint of land use planning is to resolve the conflict between competing land uses in one place that is carried out based on their economic, social and environmental suitability. Many different approaches have been designed to resolve this conflict but any of them have done. This study aimed to assess accuracy and validity compared four optimization algorithms, including linear programming (LP), simulated annealing (SA), multi-objective land use allocation (MOLA) and multidimensional choice (MDCHOICE) for land use planning in Gorgan Township. At first, land use allocation was done through the LP algorithm. Then, using achieved pixels quantity from LP optimization, the land uses was optimized in term of spatial suitability by other algorithms. The allocation maps generated through these algorithms were compared using statistical methods and landscape metrics. The map provided through MDCHOICE was used as a reference map and that generated through MOLA, SA and LP as a comparative one in a validation process. The similarities and agreements of the two maps were evaluated with various algorithms in IDRISI and MCK software. The results demonstrate that: (1) It follows that the LP algorithm to determine the optimal number of pixels to lead MDCHOICE has helped in the spatial allocation properly. (2) MOLA had a better performance for development in terms of all landscape metrics and SA algorithm was better in forestry land use allocation. Therefore the quality of spatial allocation both of them needs to be improved. (3) Although the LP algorithm is not superior to the algorithms based on the dimensions studied in this research, in terms of achieving the defined goals and the degree of deviation from the number of cells defined for each land use has the least violation. (4) The superiority of MDCHOICE over MOLA, SA and LP and applicability of various kappa statistics for comparing the maps for a more in depth analysis of agreements and disagreements.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have