Spatial Dynamics of Land Use And Land Cover Change Around The İzmir–Aydın Highway (1993–2023)

This study aims to examine the effects of increased accessibility caused by the O-31 Highway on Land Use and Land Cover (LULC) changes within a 10 km buffer zone between 1993 and 2023. Using multi-band Landsat satellite imagery, LULC classifications were performed through the Random Forest (RF) machine learning algorithm on the Google Earth Engine (GEE) platform. To enhance classification accuracy, remote sensing indices and the pan-sharpening method were integrated. Results show a significant transformation in land cover over the 30-year period. Built-up areas doubled, increasing from 8% in 1993 to 16% in 2023, while bare land expanded from 27% to 45%. Conversely, agricultural areas declined from 35% to 22%, and forest areas from 31% to 16%. Spatial analysis, including transition matrices and kernel density estimation, revealed that the most intense LULC changes were concentrated within the 1–5 km buffer zone and particularly around highway interchanges (entry/exit points). These findings highlight how accessibility-driven development tends to cluster in strategically connected peri-urban areas, aligning with broader observations in urban expansion literature while offering context-specific insights. The study underscores the limitations of current urban planning practices in managing peri-urban growth and protecting agricultural land.