Modelling spatial and temporal changes in land use/cover of Marmara Island (Türkiye) using CA-Markov model

In this study, Geographical Information Systems and Remote Sensing techniques were used to analyse the temporal spatial dynamics of land use on Marmara Island for 1991, 2001, 2011, and 2023 years representing major socio-economic transitions in Türkiye nd to model future land use for 2055 using the CA–Markov approach. NDVI values derived from multi-temporal Landsat imagery were used to assess vegetation responses to LULC change. Results indicate substantial declines in forests (38.4%), grasslands (28.3%), shrublands (54%), olive groves (15.4%) and agricultural areas (56.4%) between 1991 and 2023, while mining areas increased by 52% and are projected to expand by an additional 46% by 2055. NDVI values (–0.26 to +0.69) show pronounced vegetation loss in northern mining zones and relatively stable agricultural–forest mosaics in the south. Overall, the findings demonstrate that mining driven anthropogenic pressures are reshaping the island’s ecological structure, underscoring the need for sustainable land management.