A new method to estimate branch biomass from terrestrial laser scanning data by bridging tree structure models

Area of Science:

• Forestry science
• Remote sensing
• Ecology

Background:

• Estimating forest stand carbon budgets and crown structure relies on branch biomass data.
• Destructive biomass measurement is labor-intensive and time-consuming.
• Terrestrial laser scanning (TLS) offers non-destructive biomass estimation but faces challenges with occlusion and individual branch attribute extraction in conifers.

Purpose of the Study:

• To develop and validate a novel method (TSMtls) for non-destructive and accurate estimation of individual branch biomass.
• To integrate tree structure models with TLS data for enhanced branch attribute analysis.
• To improve the accuracy of branch biomass estimation in coniferous trees.

Main Methods:

• The TSMtls method combines tree structure models with TLS data.
• It constructs stem-taper curves from TLS data and utilizes tree models to determine branch number, basal area, and biomass at the whorl level.
• Model parameters were derived from 122 destructively measured Scots pine trees and the method was validated on six TLS-scanned and destructively measured trees.

Main Results:

• TSMtls achieved high accuracy in tree-level branch biomass estimation (CV-RMSE = 9.66%, CCC = 0.99), outperforming other TLS-based methods.
• Whorl-level individual branch attribute estimates from TSMtls were more accurate than direct TLS data analysis.
• The method demonstrated superior performance compared to existing TLS-based approaches (CV-RMSE 12.97-57.45%, CCC 0.43-0.98).

Conclusions:

• The TSMtls method shows significant promise for accurate, non-destructive branch biomass estimation.
• The approach is suitable for coniferous species and has potential for broader application across species and larger forest areas.
• This advancement aids in more precise forest inventory and carbon accounting.