Comparison of efficiency of spectral (NDWI) and SAR (GRD) method in shoreline detection: A novel method of integrating GRD and SLC products of sentinel-1 satellite

In the context of shoreline change studies for management and development purposes, it is essential to utilize historical shoreline positions. Today, various satellite methods and images are utilized; for instance, the capabilities of contemporary spectral images are well established. However, given the efficiency of Synthetic Aperture Radar (SAR) images in providing earth observations under all weather conditions, their use for identifying shoreline positions is increasingly being developed. This study presents a comparative analysis of shoreline extraction using the Ground Range Detected (GRD) and GRD-SLC (Single Look Complex) methods, alongside the spectral method as a reference. The GRD-SLC method significantly improved accuracy, particularly in estuarine areas, reducing median discrepancies from 378.0 m ( +/- 247.7) with the GRD method to 16.0 m ( +/- 26.6) for this type of shores. For steep rocky shores, the discrepancy decreased from 22.1 m ( +/- 18.4) to 8.5 m ( +/- 10.1). Conversely, the GRD method performed better for dune-backed, wide sandy, and rocky shores. Overall, the GRD-SLC method yielded a discrepancy of less than 50 m in key areas like omega-shaped beaches and cliffs, with its best performance observed in low-elevation regions (<10 m). The findings emphasize the GRD-SLC method's suitability for diverse coastal environments. Despite the limitations of the GRD method, its all-weather SAR imagery capability proves effective on a regional scale. This study highlights the need for further algorithmic advancements in SAR-based shoreline monitoring to enhance accuracy across varying coastal terrains.