Salt-marsh aboveground biomass estimation by integrating UAV LiDAR-derived canopy height and Sentinel-1/2 satellite data

Salt-marsh vegetation plays an important role in coastal wetland stability and blue carbon storage, and accurate aboveground biomass (AGB) estimation is essential for assessing wetland carbon dynamics. However, regional-scale AGB estimation based on optical remote sensing is often constrained by spectral saturation and limited sensitivity to vertical vegetation structure. To address these limitations, canopy height information derived from unmanned aerial vehicle LiDAR (UAV LiDAR) was integrated with Sentinel-2 multispectral imagery and Sentinel-1 synthetic aperture radar (SAR) data for salt-marsh AGB estimation in typical tidal flats of the Yangtze River Estuary. A canopy height model (CHM) derived from UAV LiDAR data acquired in 2023 was used to train a random forest (RF) model linking Sentinel-1/2 features with canopy height, and the trained model was applied to Sentinel-1/2 features from 2022 to 2024 to generate annual regional-scale predicted CHM maps. For each year, the predicted CHM was combined with optical features, SAR polarization features, and field-measured AGB data to develop annual AGB estimation models. The Sentinel-1/2-based CHM retrieval model achieved good performance (coefficient of determination (R2) = 0.77, root mean square error (RMSE) = 0.32 m, and relative RMSE (rRMSE) = 30.55%), demonstrating the feasibility of extending UAV LiDAR-derived canopy structure to regional scales. The multi-source model integrating optical, SAR, and CHM features achieved the best AGB estimation performance, with R2 of 0.52–0.62, RMSE of 0.35–0.62 kg/m2, and rRMSE of 31.73%–37.70%. This study provides an effective approach for regional-scale salt-marsh biomass monitoring by addressing the limited spatial coverage of UAV LiDAR and insufficient representation of vegetation vertical structure in satellite-based approaches. By integrating fine-scale canopy information from UAV LiDAR with regional and multi-temporal observations of Sentinel-1/2 imagery, the proposed framework enables continuous CHM and AGB mapping for salt-marsh ecosystem monitoring and blue carbon assessment.