What does a redwood tree look like in 3D? And why should we care? Alexander Barajas-Ritchie, a Computer Science major (SSU ’21), received a grant from Save The Redwoods League to create and refine 3D quantitative structure models (QSMs) of Coastal Redwoods (Sequoia sempervirens) to estimate their aboveground biomass.
The Issue
Coastal redwoods (Sequoia sempervirens) provide essential ecosystem services and are valuable timber species. Redwoods possess a considerable amount of aboveground biomass (AGB) which is critical for carbon accounting to quantify greenhouse gas emissions and carbon sequestration of forests. Typically, destructive harvesting of trees is used to estimate AGB, but this is time consuming and difficult.
The Solution
Alternative methods to destructive sampling, such as using light detecting and ranging (LiDAR) and quantitative structure models (QSMs), could prove to be more accurate and easier to execute than cutting down trees. In this project, we developed, optimized, and processed QSMs created from terrestrial laser scanning (TLS) data to calculate the AGB of redwood trees. We then compared QSM estimates of AGB to estimates calculated from current approaches using national and local equations created from destructive sampling.
Visualizing Redwoods in 3D!
From left to right: A redwood tree 3D image collected from the terrestrial laser scanner, the same tree with the leaves removed, and the QSM model of the tree that can be used to calculate aboveground biomass (AGB).
The Results
We were able to calculate AGB of redwoods using our approach (our results matched the results calculated from equations in the literature, as shown on the graph below) but this approach was very sensitive to specific post-processing steps (leaf separator python script and five input parameters for QSM model).
The Bottom Line
Due to the need for accurate forest carbon accounting for forest management, controlled burns, and understanding forest ecosystem health, this study is significant as it shows that we can use optimized QSMs generated from TLS data to determine the AGB of coast redwoods without destructive sampling. Importantly, the QSM fitting procedure is sensitive to certain steps, such as leaf separation, input parameters based on height class and the quality of the TLS scans, but by controlling for these issues, we can estimate the AGB of coastal redwoods in Sonoma County with great confidence.
The Acknowledgments
This study could not have been completed without funding from Save The Redwoods League, McNairs Scholars Program at Sonoma State University and CALFIRE and data collection, technical support and guidance from Brieanne Forbes and Paris Krause.
