How accurately does L band vegetation optical depth predict aboveground biomass?

Area of Science:

• Remote Sensing
• Ecology
• Biomass Estimation

Background:

• L-band Vegetation Optical Depth (L-VOD) is widely used to monitor global aboveground biomass (AGB).
• Current methods for deriving AGB from L-VOD face challenges due to methodological ambiguities and the space-for-time assumption.
• Reliability of AGB estimates is hindered by a lack of standardized protocols.

Purpose of the Study:

• To comprehensively evaluate existing methodologies for deriving AGB from L-VOD.
• To assess the impact of integrating tree cover on AGB predictions.
• To test the space-for-time assumption in AGB-L-VOD relationships.

Main Methods:

• Utilized the SMOS-ICV2 L-VOD dataset and five AGB reference datasets for evaluation.
• Assessed various fitting methods for AGB estimation from L-VOD.
• Investigated the influence of tree cover and L-VOD values on AGB prediction accuracy.
• Analyzed spatial versus temporal sensitivities of AGB to L-VOD.

Main Results:

• Existing fitting methods generally explain 69-77% of AGB spatial variation.
• Integrating tree cover significantly improves AGB predictions in areas with low to medium L-VOD.
• Methods fail to capture AGB spatial variation in dense rainforests (L-VOD > 1) where reference data also show discrepancies.
• Spatial AGB sensitivities to L-VOD are generally larger than temporal sensitivities.

Conclusions:

• Current L-VOD based AGB estimation methods have limitations, especially in dense forests.
• Improvements require integrating vegetation structural data and long-term in situ observations.
• Development of long-term field-based biomass change datasets is crucial for validating remote sensing AGB predictions.