Morphological dynamics, landscape fragmentation and climate change vulnerability of actual and predicted urban areas of Lebanon: Application of Multilayer-Perceptron Markov-Chain model

Abstract

Rapid urbanization in Lebanon poses problems of adaptation to climate change. This article aims to forecast and characterize urban growth, and to cross-analyze urban growth with anticipated temperature and precipitation information. It uses a combination of land use and land cover forecasts with the help of remote sensing and spatial analysis data to provide a comprehensive description of urban growth in Lebanon. The findings reveal that urban growth is mainly driven by road infrastructure and proximity to the coastline, rather than population growth. However, recent changes in agricultural practices are also behind urban and land cover significant evolutions. Urban patterns retain a complex shape over time, with an estimated fractal dimension of 1.7 by the year 2136. The combination of shape, aggregation and connectivity indicators for urban areas reveals a trend towards more compact and dense agglomerations, with the emergence of new centers and sub-centers. Cluster and spot analyses distinguish between clustered urban agglomerations and hotspots in inland areas, and cold spots and separated centers along the coast. Urban vulnerability to climate change is then assessed by cross-referencing information on urban morphology with expected changes in temperature and precipitation. All these elements underline the importance of taking theoretical and natural urban dynamics into account when assessing the impact of urban growth on the environment and society.