Geo-temporal analysis of land use dynamics in the Dolphin Coast, KwaZulu-Natal, South Africa, using RapidEye and PlanetScope imagery

Land use and land cover (LULC) in the Dolphin Coast of South Africa have undergone substantial transformation due to rapid urbanization and growing environmental pressures. Understanding these spatiotemporal dynamics is crucial for managing development in sensitive coastal ecosystems. This study utilized high-resolution RapidEye and PlanetScope imagery, classified through the Random Forest algorithm within the Google Earth Engine (GEE) platform, to map five LULC categories: built-up, bareland, forest, vegetation, and water. The analysis achieved overall classification accuracies ranging from 88% to 95%, confirming strong agreement between predicted and reference data. A structured change analysis was further applied to quantify land transitions by examining components of quantity, exchange, and shift. The results revealed dominant conversions from vegetation to built-up (9.0 km 2 ), bareland to built-up (3.6 km 2 ), and vegetation to forest (5.6 km 2 ). Built-up areas consistently expanded over time, while vegetation showed both gains and notable losses, reflecting intense development pressures on green spaces. The study highlights clear trajectories of landscape transformation and underscores the importance of spatially explicit change metrics for monitoring dynamic coastal environments. These findings provide valuable insights for policymakers and urban planners to design sustainable land-use frameworks, control unplanned urban expansion, and protect ecologically sensitive zones along the Dolphin Coast.