Open AccessAbove-ground biomass estimation of Eucalyptus plantation using remotely sensed data and field measurements
Author(s) -
Warakhom Wongchai,
Woravit Insuan,
Anucha Promwungkwa
Publication year - 2020
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/463/1/012042
Subject(s) - tree allometry , eucalyptus , biomass (ecology) , vegetation (pathology) , diameter at breast height , environmental science , allometry , forestry , mathematics , physical geography , statistics , geography , ecology , biology , biomass partitioning , medicine , pathology
Biomass has had an essential role in the energy sector of the world due to applications in bioenergy. Stand level biomass is frequently calculated from allometric models with field measurements, which is usually time-consuming and costly. They are limited because of the consideration of spatial pattern analysis of above-ground biomass (AGB) across the landscape. Therefore, the development of reliability and low-cost methods is necessary for AGB estimations in landscape level. This study aims to develop a model for estimating AGB for Eucalyptus plantation located in the Sahacogen Green Co., Ltd., in Lampang province, Thailand using remotely sensed data. The AGB value was coupled which calculated from field measurement (tree height, H and diameter at breast height, DBH) using the allometric equation with various vegetation indices. The 55 sample plots and 5 vegetation indices derived from Thailand Earth Observation System (THEOS) were used to develop a model for estimating AGB of Eucalyptus plantation. After discussing the results of the investigation, the Transformed Normalized Difference Vegetation Index (TNDVI) showed a robust correlation with AGB compared to other indices (r = 0.833). Based on stepwise linear regression between AGB and 5 vegetation indices demonstrated TNDVI was only selected while the other indices were eliminated because their relationship was not significant. The developed model R 2 was 0.693, adjusted R 2 was 0.684 and SEE was 12.41 Mg ha −1 . The relationship between observed AGB and predicted AGB from the THEOS model of Eucalyptus plantation with R 2 of 0.742 and RMSE of 9.63 Mg ha −1 indicated that remotely sensed data from THEOS can be useful for AGB estimation with high accuracy.