This study proposes an automated long-term shoreline extraction method based on Google Earth Engine (GEE), integrating the Automated Water Extraction Index (AWEI), Otsu thresholding, morphological operations, anomaly detection, and Savitzky-Golay smoothing to analyze the 40-year evolution (1985–2024) of the coastline at Friendship Port, Mauritania. The study area spans 20.6 km north-south and 4.2 km east-west, centered on the port. Remote sensing images were preprocessed in GEE to compute AWEI, with Otsu thresholding applied to generate water masks. Inland water bodies and small islands were removed using morphological operations, followed by coastline vectorization. Python-based post-processing detected anomalies via distance thresholds and smoothed coordinates using Savitzky-Golay filtering. The results indicate that this method enables the automated and batch extraction of coastlines. This study also compares the coastline extraction results from satellites of different spatial resolutions and those obtained through manual digitization. The findings further demonstrate that the method yields low extraction errors and that the derived coastlines show a high degree of consistency with the satellite imagery, despite the absence of tidal correction. This method provides an efficient and reliable reference for large-scale, long-term coastline monitoring and coastal dynamics analysis.