3D forest reconstruction from terrestrial laser scanning of trees for solar-induced chlorophyll fluorescence sensitivity analysis

0570665 - ÚVGZ 2024 eng A - Abstract
Janoutová, Růžena - Malenovský, Z. - Homolová, Lucie - Novotný, Jan - Navrátilová, Barbora - Pikl, Miroslav
3D forest reconstruction from terrestrial laser scanning of trees for solar-induced chlorophyll fluorescence sensitivity analysis.
[12th EARSeL Workshop on Imaging Spectroscopy in Potsdam. Potsdam, 22.06.2022-24.06.2022]
Method of presentation: Poster
URL events: https://is.earsel.org/workshop/12-IS-Potsdam2022/ 
R&D Projects: GA MŠMT(CZ) LTC20055
Institutional support: RVO:86652079
Keywords : forest reconstruction * tls * solar-induced chlorophyll fluorescence * radiative transfer modelling
OECD category: Remote sensing

This study presents a method for three-dimensional (3D) reconstruction of forest tree species that are required, for instance, for simulations of 3D canopies in radiative transfer models (RTMs). We designed and tested the method on three forest tree species of different architecture: Norway spruce (Picea abies) and European beech (Fagus sylvatica), as representatives of European production forests, and white peppermint (Eucalyptus pulchella), a common forest species of Tasmania. Each species has a very specific and distinct crown structure and foliage distribution (Janoutová et al. 2021. In Silico Plants, 3(2)).

Our algorithm for 3D model construction of a single tree is based on terrestrial laser scanning (TLS) of individual trees and ancillary field measurements of leaf angle distribution (LAD), percentage of current-year and older leaves, and other parameters that could not be derived from TLS data. The algorithm comprises of four steps: (i) segmentation of a TLS tree point cloud separating wooden parts from foliage, (ii) reconstruction of wooden parts (trunks and branches) using TLS point clouds, (iii) biologically genuine distribution of foliage within the tree crown, and (iv) separation of foliage into two age categories (for spruce trees only, Janoutová et al. 2019. Forests, 10(3)). The reconstructed 3D models of the tree species were used to build virtual forest scenes in the Discrete Anisotropic Radiative Transfer (DART) model (Gastellu-Etchegorry et al. 2017. IEEE Journal of Selected Topics in Applied Earth Obs. and Remote Sens. 10(6), Malenovský et al. 2021. Remote Sens. Environ., 263). Subsequently, canopy optical signals, specifically solar-induced chlorophyll fluorescence signal (SIF), were simulated to carry out detailed 3D sensitivity analysis that cannot be performed within field experiments.

For its correct interpretation, the remotely sensed angularly anisotropic SIF (like a top-of-canopy (TOC) SIF signal of the upcoming FLEX satellite mission) must be corrected for its interactions with canopy structures. To investigate the potential impacts of selected forest structural traits and elements, i.e. leaf area index (LAI), LAD, canopy closure (CC) and the presence/absence of wood, we carried out detailed sensitivity analysis in DART. Results showed, for instance, that removal of woody components from clumped erectophile canopies of w. peppermint with CC of 80% and 40% and LAI of 2.5 increased the far-red TOC SIF radiance by 24% and 14%, respectively, while the hemispherical SIF escape ratio (SIF TOC radiance/total SIF emission) increased by 10% and 8%, respectively, and the nadir SIF escape ratio only by 6 and 4%, respectively. Interestingly, the change in CC from 80% to 40% triggered a similar change in far-red SIF hemispherical and nadir escape ratio, i.e., a respective decrease by 8% and 6%. The elimination of woody material in the N. spruce canopy (spherical LAD, LAI = 8.5, CC ~ 80% and single generation of needle leaves) increased the far-red TOC SIF radiance almost twice (90%), whereas the hemispherical and nadir far-red SIF escape ratio increased by 42% and 40%, respectively. The greater change in the far-red TOC SIF radiance over its canopy escape ratio indicates that the effect of wood shadowing is larger than direct scattering and absorption of SIF by woody elements. Our results demonstrate a potentially significant influence of wood and other structural canopy traits that should be considered when interpreting remote sensing SIF observations.
Permanent Link: https://hdl.handle.net/11104/0341992