Application of Artificial Neural Network in Modeling and Forecasting Land Use Changes in Sardroud City (1984-2031)

Abstract

An essential step to urban planning, management and evaluation of its effects is to simulate physical development of the city. The aim of this study is to understand parameters of physical development at Sardroud city with regard to sustainable spatial development of urban issues from ecological and environmental viewpoint in the next two decades. Using Landsat 5 multi temporal satellite images and object oriented techniques, application changes of the lands in 1984-2011 period has been evaluated with emphasis on Sprawl expansion of Sardroud city. Based on the results, urban area of Sardroud which was 111.24 hectares in 1984, has reached to 528.12 ha in 2011.162.94 hectares of mentioned lands has developed on the Garden and agricultural land, which demands management of future development based on the principles of sustainable development. Therefore, effective factors of physical development in Sardroud city is classified within 14 layers, and by using Artificial Neural Network method based on LTM Model, the possibility of urban development map was prepared. After predicting the future pattern of urban development in Sardroud city, the protection of gardens and green spaces strategy in the urban development process was operated using hexagonal layout of possibility of urban development map, giving the necessary space for the development, Extraction of natural green belt with the length of 15 km applying the ban of urban development in the around of proposed green belt, maintaining the ecological reserves of Sardroud city by reducing agricultural and garden lands, and prevent connection to the metropolitan of Tabriz is emphasized.

Keywords

• sardroud city
• sprawl growth
• artificial nerual network
• land use changes
• object oriented classification