Skip to main content

Understanding forest growth dynamics using novel 3D measurements and modelling approaches

Periodic Reporting for period 1 - 3D-FOGROD (Understanding forest growth dynamics using novel 3D measurements and modelling approaches)

Reporting period: 2019-10-01 to 2021-09-30

Estimates of the global distribution of terrestrial carbon sinks and sources are highly uncertain. Constraining the inaccuracy of carbon estimates is essential to support effective forest management and future climate mitigation. A better understanding of forest growth dynamics will improve our understanding of the carbon cycle and mechanisms responsible for terrestrial carbon sources and sinks, reducing uncertainties on their magnitude and distribution. This project aims to establish an approach to more accurately estimate (above-ground) forest growth, improve our understanding of forest growth dynamics and evaluate the role of elevated CO2 levels on forest growth. This project aims to achieve this by using novel 3D terrestrial laser scanning (TLS) techniques, unique datasets and state-of-the-art modelling approaches.

3D-FOGROD uses TLS data from EucFACE (Australia) and Wytham Woods (UK). In addition to observational methods at EucFACE, simulation models are widely used to study forest growth. Second-generation vegetation models, such as the Ecosystem Demography model 2 (ED2), include demographic processes and explicitly track fine-scale ecosystem structure and function, making them an ideal tool to improve our knowledge of forest growth dynamics. ED2 will be initiated using LiDAR-derived structural metrics from Wytham Woods. ED2 allows us to explicitly represent vegetation demographic processes and use fine-scale variation in the horizontal and vertical structure and composition of forest canopies.

These are the objectives of 3D-FOREST:
(1) Accurately quantify forest growth using TLS data in a free-air CO2 enrichment experiment to determine the effect of elevated CO2 levels on forest growth
(2) Improve forest growth model dynamics using LiDAR derived forest structure to provide more accurate estimates of future carbon stocks
(3) Develop and disseminate recommendations for using terrestrial laser scanning in forest growth dynamics and carbon cycling
Objective 1 is still ongoing, Objective 2 & 3 are completed.

Objective 1:
EucFACE is a free-air CO2 enrichment experiment in a mature evergreen broadleaved forest. At the start of the project, TLS data for the six EucFACE rings was already available at three timestamps: 2012 (start of the experiment), 2015 and 2018. In March 2020, additional measurements were done, therefore adding a 4th timestamp, and extending the timeseries data to 8 years. Across the six rings, a total of 171 individual trees were segmented into individual point clouds. Tree height and vertical crown projected area were calculated per tree per timestamp. Due to some occlusion in the TLS data, we opted to use dendrometer band measurements (collected at the same day as the scanning) for the diameter measurements. This occlusion in the TLS data was due to the limited scanning setup (two locations per ring). Due to COVID-19 entry restrictions for Australia, the planned fieldwork in 2021 to collect very highly detailed scans using more scan locations could not be carried out. I therefore could not make full quantitative structure models to estimate the volume and derive biomass per tree level.
Even though limited to some extent due to the scan setup, several structural parameters could be extracted using the 3D data. Whereas the analysis of all 171 trees across all timestamps is finalized, interpretation of the results (i.e. to answer the question if enhanced levels of CO2 have an impact on the forest dynamics) is still in progress. This is due to the complexity of EUCFace, where not only enhanced CO2 is a parameter, but also other disturbances (e.g. insects, drought) occurred.

Objective 2:
We quantified the impacts of integrating structural observations of individual trees (namely tree height, leaf area, woody biomass, and crown area) derived from TLS into the state-of-the-art Ecosystem Demography model (ED2.2) at a temperate forest site Wytham Woods. We assessed the relative model sensitivity to initial conditions, allometric parameters, and canopy representation by changing them in turn from default configurations to site-specific, TLS-derived values. We show that forest demography and productivity as modelled by ED2.2 are sensitive to the imposed initial state, the model structural parameters, and the way canopy is represented. In particular, we show that:
- the imposed openness of the canopy dramatically influenced the potential vegetation, the optimal ecosystem leaf area, and the vertical distribution of light in the forest, as simulated by ED2.2;
- TLS-derived allometric parameters increased simulated leaf area index and aboveground biomass by 57 and 75%, respectively;
- the choice of model structure and allometric coefficient both significantly impacted the optimal set of parameters necessary to reproduce eddy covariance flux data.

Objective 3:
The fellow participated in an interdisciplinary workshop on ‘Degradation of Tropical Forests: observations, modelling and socio-environmental implications’ in Manaus, Brazil. Outcomes of this workshop are currently being prepared for publication.
In 2019, the fellow was the main organiser of the 4th Scientific Meeting on TLS in Forest Ecology (hosted by CAVElab at Ghent University). This interdisciplinary two-day interactive conference attracted ~80 participants globally and resulted in a review & perspective paper.
The fellow was the lead-author of the chapter "Terrestrial Laser Scanning & UAVs" in Aboveground Woody Biomass Product Validation Good Practices Protocol. Version 1.0. by Duncanson et al. (2021).
This is the first (to the fellow's knowledge) detailed 4D timeseries at individual tree-level in forest ecosystems that covers a significant amount of years. Whereas interpretation of the results is still ongoing, it is expected to answer the question if enhanced levels of CO2 have an impact on the forest dynamics. Over the years, an increased liana load is observed, and his TLS timeseries offered a unique opportunity to track the impact of lianas on tree structure (a new collaboration with the University of Nottingham). The newly collected 2020 TLS data at EucFACE will be made available together with the data collected in 2012, 2015 and 2018 in a public repository with the publication related to objective 1.
Furthermore, we conclude that integrating vegetation structure information derived from TLS can inform Terrestrial Biosphere Models (TBMs) on the most adequate model structure, constrain critical parameters, and prescribe representative initial conditions. This work also confirms the need for simultaneous observations of plant traits, structure and state variables if we seek to improve the robustness of TBMs and reduce their overall uncertainties.

In 2020, the fellow collaborated with artists James McGrath and Gary Sinclair who created ‘nature as data’, an exhibition of vulnerable trees to bring awareness of the impact humans are having on these forest ecosystems. Using TLS data as input, a mixed media installation was created, that explores the visual and emotional aspects of scientific data on the environment and combined 3D TLS and ecoacoustic data (https://www.james-mcgrath.com/mcgrath_wordpress_portfolio/ghost-trees) and already featured in Sydney and Melbourne.
TLS fieldwork at EUCFace
Illustration of growth and increase in liana load in ring 5: 2012 vs 2020.