Artificial reefs (ARs) are increasingly deployed to mitigate the degradation of natural marine ecosystems; however, quantitatively resolving early-stage, multi-species colonization at millimetre–centimetre scales remain challenging under field conditions. This study evaluates an uncertainty-aware workflow integrating Structure-from-Motion (SfM) underwater photogrammetry and Multiscale Model-to-Model Cloud Comparison (M3C2) for multi-temporal structural-change detection on two purpose-built AR modules (1.0 × 1.0 m) deployed at about 10 m depth off Silifke, Mersin. Six underwater surveys were conducted over one year (Nov 2023–Dec 2024), complemented by laboratory geodetic control for independent validation. Dense point clouds were reconstructed from overlapping imagery and assessed using total-station check points (CPs), yielding millimetric positional accuracy (Three-Dimensional Root Mean Square Error (3D RMSE): 1.56–2.66 mm; Epoch 2 excluded due to sand burial of CPs). Geometric consistency between the laboratory baseline and the post-deployment underwater baseline was further supported by Cloud-to-Cloud (C2C) comparison (means: 0.7–1.0 mm). M3C2 analyses against the underwater baseline (Epoch 1) produced Level of Detection at 95% confidence (LoD95%) thresholds of 2.06–3.14 mm (AR1) and 2.08–2.91 mm (AR2) and revealed statistically significant positive surface displacements consistent with progressive colonization, with cumulative surface growth of about 23 mm (AR1) and about 30 mm (AR2) over the monitoring period. The results demonstrate that SfM-based underwater photogrammetry coupled with M3C2 provides a non-invasive, repeatable structural-change metric that complements (rather than replaces) conventional in-water ecological surveys for AR monitoring.
Hamal et al. (Thu,) studied this question.