Abstract
Tidal range, diverse land-cover conditions, and complex intertidal topography make accurately monitoring estuarine shoreline change difficult. This study extends the COASTGUARD workflow from traditional open coastline applications to complex, macrotidal estuarine environments, demonstrated through a case study in the Musquash Marine Protected Area. The revised pipeline introduces three key contributions: habitat-aware classification tailored to estuarine settings, a buffer-area strategy for tidal flat shorelines, and tidal correction based on DEM-derived slope estimation. High-resolution SuperDove imagery was used for habitat classification using a three-layer MLP with 91.94% accuracy validated against existing habitat maps. Slopes were derived through calculations of shoreline position using a 1 m LiDAR derived DEM. Water level was calculated through applying the Piecewise Cubic Hermite Interpolation on a seven-day moving window. Images captured during high tide had a 150 m buffer applied to the reference shoreline , reducing misclassification from offshore noise . These contributions to the COASTGUARD framework enhance shoreline tracking in a macrotidal environment of beaches, salt marshes, and tidal flats in close proximity. This introduces an accurate long term and event-based monitoring strategy for environments hosting across different tidal zones, advancing applied scientific research and management applications.