Coverage-aware adaptive partitioning for MBES bathymetric survey-line planning over complex seabed terrain

Multibeam echosounder (MBES) bathymetric surveys must satisfy strict coverage requirements even where swath width varies with terrain. This paper proposes an offline coverage-aware adaptive partitioning framework for MBES survey-line planning over complex seabed terrain. The framework constructs a gridded terrain model, derives a prior coverage-count matrix to represent local coverage opportunity, combines spatial position, depth, coverage count, and terrain gradient during partition adjustment, and generates survey lines within the resulting partitions. All main comparison groups are evaluated with the same fine-grid coverage model. In an anonymous real seabed case of about 7408 × 9260 m, the reported 9-partition E4 configuration increases sample-based geometric coverage from 83.59% under fixed spacing to 99.12%, reducing missed coverage from 16.41% to 0.88%. Relative to static partitioning, E4 raises coverage from 95.22% to 99.12% and shortens the total survey-line length by 3.28 km. The E3 ablation is more efficient in line number, length, and redundancy, so the results indicate a coverage–efficiency trade-off rather than uniform dominance by E4. A literature-adapted seabed planner achieved slightly higher coverage (99.43%) but higher redundancy (32.92%) at similar survey-line length. Supplementary checks also examine parameter sensitivity, alternative baselines, and a public H11879 transfer case. The reported rates are geometric coverage indicators under the adopted swath model and do not represent full hydrographic survey-quality certification.