Comparison of line intercept transect method, underwater photo transect, underwater 3D photo, drone, and satellite imagery in coral reef survey

ACCURATE CORAL Reef monitoring is achieved using methods that capture ecological, structural, and spatial variability across scales. This study aims to quantitatively compare the Line Intercept Transect (LIT), Underwater Photo Transect (UPT), underwater 3D photogrammetry, Unmanned Aerial Vehicles (UAVs), and Sentinel-2 imagery. Field surveys were conducted at Gili Labak and Gili Genting Islands, Indonesia. The coral cover estimates differed significantly between LIT and UPT (p = 0.049), with ranges of 13.31–64.9% (LIT) and 14.1–72.0% (UPT) at Gili Genting, as well as 46.0–72.3% (LIT) and 45.5–84.5% (UPT) at Gili Labak. Underwater 3D photogrammetry showed high predictive accuracy for biomass estimation (R 2 = 0.8869–0.9149; RMSE 17.7–23.9%). UAV-based classification achieved absolute reliability (overall accuracy 83%; Kappa 0.79) at extremely high spatial resolution (2.08 cm/pixel). Meanwhile, Sentinel-2 classification showed moderate performance (overall accuracy 63%; Kappa 0.49) due to coarser resolution (10–20 m), with coral cover estimates of 20.5–27.6%. Underwater 3D photogrammetry provided the highest accuracy for structural and biomass estimation, with UAVs offering optimal spatial resolution for site-scale mapping. Simultaneously, satellite imagery was the most suitable for large-scale monitoring despite its lower accuracy. The novelty of this study centered on the quantitative integration of multiscale methods that connected in situ and remote sensing data.