Comparison of Ecosystem- Based Land Allocation Using Genetic Algorithm and MOLA

Land use planning is a broad term that can be applied to different processes related to management of land use. One of the most complex tasks in this process is allocating land use categories to spatial units, resulting in a land use zoning map. Specifying the appropriate land unit for land allocation is one of the typical issues that are mainly divided into two categories including cell (such as MOLA approach) and polygon units (such as map overlay and systemic analysis). Optimization algorithms are the part of land allocation methods that have both Multi-Objective approach and cell or polygon structures. In this study, the Genetic Algorithm (GA) is used for land allocation based on suitability and landscape metrics in Gorgan Township. Moreover, in an innovative approach, object-oriented classification (based on environmental parameters) was used to create ecosystem units and land allocation was applied to these units. The results showed that land allocation through ecosystem-based Genetic Algorithm leads to a significant improvement of landscape metrics in comparison with MOLA. The genetic algorithm approach improved four landscape metrics including number of patches, contiguity, and effective mesh size and cohesion index. In this process, land use was allocated to homogeneous units in terms of ecological resources. Thus, diversity was minimum in the environmental units considered for the land allocation processes.