Modelling and simulation of urban dynamics

This study aims to establish a preliminary framework for modeling and simulation of urban dynamics at the Istanbul metropolitan area. The past, current and possible future land-use dynamics is described and analyzed by means of spatial information sciences. For simulation, two different stochastic algorithms including Logistic Regression and Cellular Automata based Markov models are used. The core data set is satellite images. In order to achieve spatially consistent datasets that cover large areas with both high detail and high temporal frequency, including historical time series, remote sensing is used. The core data set for the land-cover information within this study consists of satellite images, which are acquired by the Landsat satellites in 1984, 1995 and 2006. The land-use information for the year 2017 is predicted. The differences between observation and simulation are presented. Advantages and disadvantages of both algorithms were discussed. According to the achieved results, the proposed framework might be used for aiding authorities via simulation results of their taken decisions.