Accurate quantification of forest carbon stocks depends on reliable measurement of forest structure: canopy height, vertical foliage distribution, and ground elevation beneath canopy. Forest carbon is not directly observable from space; it must be inferred from structural proxies such as canopy height and biomass structure. Errors in any component propagate into above-ground biomass (AGB) and carbon estimates. Current global remote sensing approaches still exhibit substantial uncertainty.
Multiple studies demonstrate that uncertainty in mapped forest biomass often remains at tens of percent, even when using advanced remote sensing and modelling techniques. Global goods like pantropical AGB maps have typical pixel-scale uncertainties on the order of ~30% and are subject to substantial regional variation.
This persistent uncertainty underscores the need for more structurally informative remote sensing systems, because large errors can undermine decision-grade carbon accounting, emissions reporting, and carbon market verification.
Optical imagery (e.g., multispectral satellite data) has enabled large-scale vegetation monitoring, but tends to saturate at higher biomass levels and lacks direct vertical structural information.
Active radar systems (SAR) penetrate canopy to an extent and have been used for AGB estimation. However, the radar backscatter signal is influenced by moisture, structure, and terrain effects, which complicates consistent biomass retrieval across biomes without extensive calibration. Although SAR enhances structural sensitivity over optical alone, uncertainty remains when extrapolating across different forest types and conditions.
Spaceborne LiDAR missions such as NASA’s Global Ecosystem Dynamics Investigation (GEDI) have transformed vertical canopy measurement at large scales by providing 3D structure information through waveform LiDAR.
However, GEDI and similar space missions operate with a sparse footprint sampling pattern (e.g., ~25 m footprints along orbital tracks) that requires interpolation or integration with other sensors to produce wall-to-wall maps. This sampling sparsity, while extremely valuable for structural information, limits continuous high-resolution mapping without ancillary data.
https://www.youtube.com/watch?v=qpzFn5bqhl4
Airborne laser scanning (ALS) is widely used for detailed forest structure mapping and is regarded as a gold standard in biomass estimation because it provides dense 3D point clouds (better than 1 point per square metre) over target areas.
This high accuracy arises from three critical capabilities:
These capabilities collectively reduce uncertainty in height and proxy variables used for biomass estimation.
Accurate determination of ground elevation is essential because canopy height - the difference between top-of-canopy and ground - is the principal predictor of biomass. Any systematic bias in ground elevation retrieval directly biases canopy height and, by extension, biomass estimates.
High spatial resolution increases the likelihood of sampling small canopy gaps and detecting ground returns, particularly under dense forest cover, whereas larger footprints average returns over mixed canopy and terrain and can obscure true ground surfaces.
Forests exhibit spatial heterogeneity at fine scales: crowns, gaps, understory variation, and microscale terrain. High horizontal resolution (~1 m) resolves individual crowns and small-scale gaps, enabling more physically meaningful measurements of structure and greater accuracy in derived carbon products.
By contrast, coarser resolutions (e.g., ≥10–30 m) blend heterogeneous structure into averaged signals, reducing vertical detail and increasing dependence on statistical models to estimate biomass.
For remote sensing–based biomass estimation capable of supporting ≤10% uncertainty globally, the measurement system should meet the following minimum requirements:
Airborne LiDAR meets these structural requirements locally and empirically demonstrates high accuracy. The following sections will examine the limitations of aerial LiDAR, and why spaceborne LiDAR architectures are necessary.