Estimation of the aboveground carbon stocks based on tree species identification in Saihanba plantation forest
The expansion of plantation forests has been identified as a significant potential strategy for natural initiatives aimed at mitigating the ongoing global climate change. Consequently, the rapid and accurate estimation of aboveground carbon stocks in plantation forests is a crucial prerequisite for...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-04-01
|
| Series: | Ecological Indicators |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1470160X25003000 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The expansion of plantation forests has been identified as a significant potential strategy for natural initiatives aimed at mitigating the ongoing global climate change. Consequently, the rapid and accurate estimation of aboveground carbon stocks in plantation forests is a crucial prerequisite for regional carbon balance assessment and the formulation of global climate change response strategies. In this study, based on the multi-temporal Sentinel-2 remote sensing image data, five classification schemes were devised to identify the dominant tree species of all forest patches or sub-compartments in the Saihanba Mechanical Forestry using the Random Forest algorithm. Then, the sub-compartments were classified by dominant tree species, and the aboveground carbon stocks of major five types with more than 500 sub-compartments were estimated through a spatial gridded method. The results were shown that: 1) The identification effect of Scheme IV, as ascertained by screening three types of effective feature vectors based on the random forest algorithm, was the most effective, with an overall accuracy (OA) and kappa coefficient of 89.7% and 0.863, respectively. Moreover, the identification effect of Larch (Larix principis-rupprechtii Mayr) was the most optimal in all schemes, with the highest PA of 92.9%. The most obvious enhancement in the identification accuracy of Spruce (Picea asperata Mast.) was evident, with a notable increase in PA from 38.0% to 55.7%, signifying a substantial 17.7% rise; 2) The dominant tree species in Saihanba exhibited distinct spatial patterns. White Birch (Betula platyphylla Sukaczev) was primarily distributed in the Yinhe Forest in the east, Mongolica (Quercus mongolica Fisch) was predominantly concentrated in the Dahuanqi Forest in the south, Pinus (Pinus sylvestris var. mongholica Litv.) was mainly distributed in the west, the Sandaohekou Forest and the Qiancengban Forest, and Larch were found throughout the entire region. 3) In conjunction with the findings of tree species identification via the gridding estimation approach, it became evident that the aggregate aboveground carbon stocks of each dominant tree species within the Saihanba Mechanical Forestry region amounts to 3313.28 thousand tons. Furthermore, the discrepancy between this estimation and the direct measurement outcomes of the forest management inventory (FMI) was minimal, exhibiting a relative error of only −5.2%, estimation precision of 94.8%. This study would provide a critical methodological advancement for regional/local-scale carbon stocks monitoring in plantation ecosystems, with both scientific and practical implications for global climate governance. |
|---|---|
| ISSN: | 1470-160X |