An integrative modelling approach to analyse landscape dynamics through intensity analysis and cellular automata-markov chain model

Abstract

The goal of this study is offer a deep understanding of the landscape dynamics in the Gorgan Township, the Golestan Province, Iran. Landsat satellite imagery of two different time thresholds, i.e. the years 1992 and 2011, was acquired from the US Geological Survey database and the changes were quantified for the Gorgan area covering a 19-year time span. Furthermore, an integrated Cellular Automata-Markov Chain (CA-MC) model was applied to predict future changes up to the year 2030. We used the intensity analysis method to compare the historical dynamics of different land categories at multiple levels. The results indicated that during the 19 years, the built-up and forest areas increased by 2.33% and 0.27%, respectively, while agriculture and remnant vegetation decreased by 2.43% and 0.24%, respectively. The CA-MC model illustrated that in the following 19 years, the built-up areas could increase by 2.45%. An intensity analysis revealed that forest gains and losses were dormant while remnant vegetation gains and losses were active. The built-up area's gains and water bodies' losses were active and stationary during both time intervals. The transitions from water bodies and remnant vegetation to agriculture were regularly targeting and stationary, while the transition from forest to agriculture was regularly avoiding and stationary. Our findings also indicated a heavy systematic transition from agriculture to built-up areas. Regarding the increasing population growth and urbanisation in the region, the outcomes of this study can help make informed decisions for the management and protection of natural resources in the study area.