Modeling of forest canopy density confusion in environmental assessment using artificial neural network

Environmental Impact Assessment (EIA) is well-known as a basic tool for environmental management and sustainable development. However, modelling approaches are generally preferred when quantitative entities are required for decision-making. The purpose of this study was to test artificial neural network incorporating ecosystem components, forest management activities and the forest canopy density confusion. The study area embraced three districts of Patom, Namkhaneh and Gorazbon within Khyroud research and educational forest of University of Tehran. Land Management Units were formed using available ecological databases and GIS. Based on qualitative and quantitative measures of ecological condition and human activities, the forest canopy density was simulated using artificial neural networks in Neuro Solutions ver. 5 software. Multilayer Perceptron network with one hidden layer and four neurons created the best function for optimizing topology with highest coefficient of determination ~ 0.9864. The results of sensitivity analysis revealed that human activities like the cattle density in land unit (ha), ecological and natural factors such as the average diameter of forest type trees and soil depth are associated with the highest effects on forest canopy density. As a conclusion, the impact assessment of implemented projects could offer an improved solution in decision making for similar projects across similar locations.