Answers are grounded in this paper and cite their sources.
Coastal restoration programmes must decide which projects to inspect within finite capacity, yet supervisory records leave many assessment questions unresolved. We formulate inspection allocation as a lexicographic integer programme on a source-linked project-domain ledger: the selected projects first minimize the widest residual gap in unresolved records across assessment domains and then maximize the number of questions addressed. The East China application codes 72 supervisory records from 18 projects as documented concern, scoped negative or unresolved under an explicit codebook, with delivery records and ecological entries from five separate projects as context. Unresolved records (39 of 72) concentrated in the quality/maintenance and implementation-scheme domains, for which reports seldom recorded explicit statements. At six equal project slots, the exact allocation targeted 19 questions versus 13 expected under uniform random selection and left a widest residual interval of 44.4 percentage points, compared with 54.4 for random selection and 50.0 for five other maximum-volume sets. Quadratic-balance, gap-weighted and value-of-information minimax objectives selected the same projects, whereas risk-based and concern-first rules served corrective follow-up. A transparent sequential domain-balance rule reproduced the exact optimum at every budget on this ledger and the exact widest gap in 99.2% of 24,000 simulated cases. In the separate Beilun case, surveyed in 2021 and 2025 without contemporaneous reference sites, benthic biomass increased at 12 of 15 matched stations and one station dominated the mean change; the station register locates where season-matched repeat observation is needed. The workflow yields an auditable inspection plan that is updated as evidence arrives.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: