The Applicability of the Extended Markov Chain Model to the Land Use Dynamics in Lebanon

This study validates the applicability of the extended Markov chain modeling for predicting the dynamics of land use in Lebanon by using (i) the current available population density maps, (ii) Landsat satellite images, (iii) ArcGIS software and (iv) TerrSet software. An extended Markov chain model was developed to predict future land use dynamics by integrating the built-up area into three categories as per the development density based on population density maps for years 2000, 2009 and 2018. Land use classification maps were generated by integrating medium-resolution Landsat images of 30 meters into ArcGIS. The population density maps were combined with the classified land use maps in order to divide the built-up land use areas into three categories: high-, medium- and low-density urban development. The observed transition matrices were computed using TerrSet software. A Markov chain projection of land use for year 2018 was generated based on the actual transition matrix and compared to the observed land use. A Pearson's chi-square test was conducted to validate the proposed model. Then, a projection of land use for 2036 was conducted based on the actual transition matrix between 2000 and 2018. Finally, insights and recommendations were proposed to address future Lebanese land use dynamics.