Answers are grounded in this paper and cite their sources.
Accurate estimation of aboveground biomass (AGB) in carbon-dense mangrove forests is central to global carbon accounting. Light Detection and Ranging (LiDAR) provides direct information on three-dimensional canopy structure and is therefore well suited for estimating AGB in structurally complex forests. However, LiDAR–AGB models remain constrained by limited ground truthing in intertidal zones and by the difficulty of separating the effects of spatial scale, stand development, and metric selection. To disentangle these factors, we designed a controlled virtual forest experiment that coupled an individual-based mangrove model with a LiDAR simulator. We analyzed plot-level LiDAR–AGB relationships from over 57 million cumulative simulated tree records with complete reference data on stem diameter, height, species identity, and biomass. This dataset allowed us to evaluate LiDAR–AGB models across plot sizes, canopy height model (CHM) pixel sizes, LiDAR metrics, and successional states. Spatial aggregation strongly improved predictive accuracy, with 50 × 50 m plots reducing the prediction error of LiDAR–AGB power-law models to below 5%. This relationship also varied with successional state. In recovering stands, canopy height dynamics and biomass accumulation remained closely coupled, producing a near-linear LiDAR–AGB relationship across spatial scales. In mature stands, canopy height became increasingly saturated, whereas biomass continued to accumulate through stand-level basal area growth, resulting in a more sub-linear relationship as plot size increased. By linking individual-level forest structure and biomass to emergent LiDAR signals, our virtual modelling approach provides a mechanistic complement to empirical calibration and supports more robust mangrove carbon monitoring across spatial scales.
No takes yet. Share an insight, caveat, or question.
Liu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: