From mangrove gain to persistent establishment: a harmonised Landsat framework for biomass-carbon proxy change and restoration screening in the United Arab Emirates, 1990–2023

Mangrove expansion is increasingly central to climate and restoration policy in the United Arab Emirates (UAE), but endpoint gain alone does not show whether mapped establishment persists or where biomass accumulation is concentrated. We developed a harmonised 30 m Landsat workflow for 1990, c.2000, 2013 and 2023 that links mangrove extent to biomass-carbon proxy change, multi-epoch gain persistence and spatial screening. Random Forest extent maps were assessed using buffered spatial cross-validation and Global Mangrove Watch (GMW) benchmarking; the 2023 map showed a Dice agreement of 0.770 with GMW v4.0.19 and a 30 m tolerance harmonic agreement of 0.881. Mapped mangrove area increased from 46.5 to 84.8 km 2 , while biomass carbon increased from 6,456 to 40,097 Mg C, yielding a net change of 33,641 Mg C (123,463 Mg CO 2 e). Endpoint gain covered 6,412.68 ha and contained 100.5 thousand Mg of positive biomass ΔCO 2 e. Persistent mapped gain, comprising long-term and decadal persistent-gain classes, covered 4,387.77 ha, representing 68.4% of gain area but 76.2% of the positive gain-associated biomass signal. Persistence accounted for 90.7% of the ΔCO 2 e signal in the original top-10 hotspots, and nine of these cells remained in the persistence-ranked top 10. Only 66 persistent patches ≥10 ha retained 69.0% of the persistence-screened signal. The workflow moves beyond extent change by distinguishing endpoint gain from persistent mapped establishment and identifying candidate areas for restoration-outcome monitoring, field verification and blue-carbon pre-feasibility. Biomass estimates remain proxy-based and are not credit-issuance calculations.