Integrating Airborne LiDAR and Terrestrial Laser Scanner forest parameters for accurate above-ground biomass/carbon estimation in Ayer Hitam tropical forest, Malaysia

Muluken N. Bazezew (Corresponding Author), Yousif A. Hussin, E.H. Kloosterman

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Parameters of individual trees can be measured from LiDAR data provided that the laser points are dense enough to distinguish tree crowns. Retrieving tree parameters for above-ground biomass (AGB) valuation of the complex biophysical tropical forests using LiDAR technology is a major undertaking, and yet needs vital effort. Integration of Airborne LiDAR Scanner (ALS) and Terrestrial Laser Scanner (TLS) data for estimation of tree AGB at a single-tree level has been investigated in part of the tropical forest of Malaysia. According to the complete tree-crown detection potential of ALS and TLS, the forest canopy was cross-sectioned into upper and lower canopy layers. In a first step, multiresolution segmentation of the ALS canopy height model (CHM) was deployed to delineate upper canopy tree crowns. Results showed a 73% segmentation accuracy and permissible to detect 57% of field-measured trees. Two-way tree height validations were executed, viz. ALS-based upper and TLS-based lower canopy tree heights. The root mean square error (RMSE) for upper canopy trees height was 3.24 m (20.18%), and the bias was –1.20 m (–7.45%). For lower canopy trees height, RMSE of 1.45 m (14.77%) and bias of 0.42 m (4.29%) were obtained. In a second step, diameter at breast height (DBH) of individual tree stems detected from TLS data was measured. The RMSE obtained was 1.30 cm (6.52%), which was as nearly accurate as manually measured-DBH. In a third step, ALS-detected trees were co-registered and linked with the corresponding tree stems detected by TLS for DBH use. Lastly, an empirical regression model was developed for AGB estimated from a field-based method using an independent variable derived from ALS and TLS data. The result suggests that traditional field-methods underestimate AGB or carbon with the bias –0.289 (–3.53%) Mg, according for approximately 11%. Conversely, integrative use of ALS and TLS can enhance the capability of estimating more accurately AGB or carbon stock of the tropical forests.
Original languageEnglish
Pages (from-to)638-652
Number of pages15
JournalInternational Journal of Applied Earth Observation and Geoinformation
Volume73
Early online date8 Sep 2018
DOIs
Publication statusPublished - 1 Dec 2018

Fingerprint

aboveground biomass
scanner
tropical forest
Biomass
laser
Carbon
Lasers
carbon
Mean square error
canopy
parameter
segmentation
stem
field method
forest canopy
valuation

Keywords

  • ITC-ISI-JOURNAL-ARTICLE

Cite this

@article{6769cce623c74476b16e34188e1b1c40,
title = "Integrating Airborne LiDAR and Terrestrial Laser Scanner forest parameters for accurate above-ground biomass/carbon estimation in Ayer Hitam tropical forest, Malaysia",
abstract = "Parameters of individual trees can be measured from LiDAR data provided that the laser points are dense enough to distinguish tree crowns. Retrieving tree parameters for above-ground biomass (AGB) valuation of the complex biophysical tropical forests using LiDAR technology is a major undertaking, and yet needs vital effort. Integration of Airborne LiDAR Scanner (ALS) and Terrestrial Laser Scanner (TLS) data for estimation of tree AGB at a single-tree level has been investigated in part of the tropical forest of Malaysia. According to the complete tree-crown detection potential of ALS and TLS, the forest canopy was cross-sectioned into upper and lower canopy layers. In a first step, multiresolution segmentation of the ALS canopy height model (CHM) was deployed to delineate upper canopy tree crowns. Results showed a 73{\%} segmentation accuracy and permissible to detect 57{\%} of field-measured trees. Two-way tree height validations were executed, viz. ALS-based upper and TLS-based lower canopy tree heights. The root mean square error (RMSE) for upper canopy trees height was 3.24 m (20.18{\%}), and the bias was –1.20 m (–7.45{\%}). For lower canopy trees height, RMSE of 1.45 m (14.77{\%}) and bias of 0.42 m (4.29{\%}) were obtained. In a second step, diameter at breast height (DBH) of individual tree stems detected from TLS data was measured. The RMSE obtained was 1.30 cm (6.52{\%}), which was as nearly accurate as manually measured-DBH. In a third step, ALS-detected trees were co-registered and linked with the corresponding tree stems detected by TLS for DBH use. Lastly, an empirical regression model was developed for AGB estimated from a field-based method using an independent variable derived from ALS and TLS data. The result suggests that traditional field-methods underestimate AGB or carbon with the bias –0.289 (–3.53{\%}) Mg, according for approximately 11{\%}. Conversely, integrative use of ALS and TLS can enhance the capability of estimating more accurately AGB or carbon stock of the tropical forests.",
keywords = "ITC-ISI-JOURNAL-ARTICLE",
author = "Bazezew, {Muluken N.} and Hussin, {Yousif A.} and E.H. Kloosterman",
year = "2018",
month = "12",
day = "1",
doi = "10.1016/j.jag.2018.07.026",
language = "English",
volume = "73",
pages = "638--652",
journal = "International Journal of Applied Earth Observation and Geoinformation (JAG)",
issn = "1569-8432",
publisher = "Elsevier",

}

TY - JOUR

T1 - Integrating Airborne LiDAR and Terrestrial Laser Scanner forest parameters for accurate above-ground biomass/carbon estimation in Ayer Hitam tropical forest, Malaysia

AU - Bazezew, Muluken N.

AU - Hussin, Yousif A.

AU - Kloosterman, E.H.

PY - 2018/12/1

Y1 - 2018/12/1

N2 - Parameters of individual trees can be measured from LiDAR data provided that the laser points are dense enough to distinguish tree crowns. Retrieving tree parameters for above-ground biomass (AGB) valuation of the complex biophysical tropical forests using LiDAR technology is a major undertaking, and yet needs vital effort. Integration of Airborne LiDAR Scanner (ALS) and Terrestrial Laser Scanner (TLS) data for estimation of tree AGB at a single-tree level has been investigated in part of the tropical forest of Malaysia. According to the complete tree-crown detection potential of ALS and TLS, the forest canopy was cross-sectioned into upper and lower canopy layers. In a first step, multiresolution segmentation of the ALS canopy height model (CHM) was deployed to delineate upper canopy tree crowns. Results showed a 73% segmentation accuracy and permissible to detect 57% of field-measured trees. Two-way tree height validations were executed, viz. ALS-based upper and TLS-based lower canopy tree heights. The root mean square error (RMSE) for upper canopy trees height was 3.24 m (20.18%), and the bias was –1.20 m (–7.45%). For lower canopy trees height, RMSE of 1.45 m (14.77%) and bias of 0.42 m (4.29%) were obtained. In a second step, diameter at breast height (DBH) of individual tree stems detected from TLS data was measured. The RMSE obtained was 1.30 cm (6.52%), which was as nearly accurate as manually measured-DBH. In a third step, ALS-detected trees were co-registered and linked with the corresponding tree stems detected by TLS for DBH use. Lastly, an empirical regression model was developed for AGB estimated from a field-based method using an independent variable derived from ALS and TLS data. The result suggests that traditional field-methods underestimate AGB or carbon with the bias –0.289 (–3.53%) Mg, according for approximately 11%. Conversely, integrative use of ALS and TLS can enhance the capability of estimating more accurately AGB or carbon stock of the tropical forests.

AB - Parameters of individual trees can be measured from LiDAR data provided that the laser points are dense enough to distinguish tree crowns. Retrieving tree parameters for above-ground biomass (AGB) valuation of the complex biophysical tropical forests using LiDAR technology is a major undertaking, and yet needs vital effort. Integration of Airborne LiDAR Scanner (ALS) and Terrestrial Laser Scanner (TLS) data for estimation of tree AGB at a single-tree level has been investigated in part of the tropical forest of Malaysia. According to the complete tree-crown detection potential of ALS and TLS, the forest canopy was cross-sectioned into upper and lower canopy layers. In a first step, multiresolution segmentation of the ALS canopy height model (CHM) was deployed to delineate upper canopy tree crowns. Results showed a 73% segmentation accuracy and permissible to detect 57% of field-measured trees. Two-way tree height validations were executed, viz. ALS-based upper and TLS-based lower canopy tree heights. The root mean square error (RMSE) for upper canopy trees height was 3.24 m (20.18%), and the bias was –1.20 m (–7.45%). For lower canopy trees height, RMSE of 1.45 m (14.77%) and bias of 0.42 m (4.29%) were obtained. In a second step, diameter at breast height (DBH) of individual tree stems detected from TLS data was measured. The RMSE obtained was 1.30 cm (6.52%), which was as nearly accurate as manually measured-DBH. In a third step, ALS-detected trees were co-registered and linked with the corresponding tree stems detected by TLS for DBH use. Lastly, an empirical regression model was developed for AGB estimated from a field-based method using an independent variable derived from ALS and TLS data. The result suggests that traditional field-methods underestimate AGB or carbon with the bias –0.289 (–3.53%) Mg, according for approximately 11%. Conversely, integrative use of ALS and TLS can enhance the capability of estimating more accurately AGB or carbon stock of the tropical forests.

KW - ITC-ISI-JOURNAL-ARTICLE

UR - https://ezproxy2.utwente.nl/login?url=https://library.itc.utwente.nl/login/201/isi/hussin_int.pdf

U2 - 10.1016/j.jag.2018.07.026

DO - 10.1016/j.jag.2018.07.026

M3 - Article

VL - 73

SP - 638

EP - 652

JO - International Journal of Applied Earth Observation and Geoinformation (JAG)

JF - International Journal of Applied Earth Observation and Geoinformation (JAG)

SN - 1569-8432

ER -