Assessing mangrove resilience in arid coasts: A case study from Egypt's Red Sea (2017-2024)

This dataset supports a spatiotemporal assessment of mangrove distribution and its environmental drivers along the Egyptian Red Sea coast between 2017 and 2024. The study area extends from Hurghada in the north to Shalaten in the south, covering multiple mangrove stands subjected to varying environmental and anthropogenic conditions. The dataset includes final mangrove classification outputs, environmental raster layers, accuracy assessment data, and analysis scripts, enabling transparency and reproducibility of the research workflow. Mangrove extent was derived from Sentinel-2 imagery using a combination of vegetation indices: NDVI >0.2 MSAVI2 (threshold > 0.3) Water areas were masked out using the NDWI index to reduce commission errors. The combined NDVI/MSAVI mask, with water masked by NDWI, was used to produce the final binary mangrove masks, which are archived here. Environmental drivers archived as GeoTIFF files include: Rainfall Sea Surface Temperature (SST) Sea Surface Salinity (SSS) These variables were aggregated consistently across locations and years to examine their relationship with mangrove area dynamics. For each study location, the dataset contains: Yearly binary mangrove masks Accuracy assessment point layers generated using a stratified random sampling approach, with balanced samples inside and outside mangrove extents Statistical analysis was conducted using a mixed-effects regression model, treating location as a random effect and environmental variables as fixed effects. The analysis code is provided as a documented text file describing the full modeling procedure. The dataset is organized into clearly structured folders separating environmental variables, mangrove classification results, accuracy assessment data, scripts, and documentation.