Tree density has been underestimated in the mountainous regions of Northeast China
Previous attempts to quantify tree density have often underestimated the numbers of trees in mountainous regions with complex terrain. We surveyed trees with a diameter at breast height (DBH) of ≥10 cm across 1,926 plots. By utilizing recursive feature elimination (RFE), we identified six key variab...
Saved in:
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-07-01
|
| Series: | Ecological Indicators |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1470160X25005850 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Previous attempts to quantify tree density have often underestimated the numbers of trees in mountainous regions with complex terrain. We surveyed trees with a diameter at breast height (DBH) of ≥10 cm across 1,926 plots. By utilizing recursive feature elimination (RFE), we identified six key variables for our meta-learner in the stacking process, including the soil silt content, soil clay content, elevation, Normalized Difference Vegetation Index (NDVI), precipitation in the wettest month, and precipitation in the coldest quarter, all of which were found to influence tree density. We developed a stacking ensemble learning algorithm, which ultimately generated a tree density map with a spatial resolution of 30 m for the mountainous regions of Northeast China. The estimated tree count is approximately 27.497 billion. Compared to global tree density datasets, our approach increased R2 to 0.454, while root mean square error (RMSE) and bias improved by 47.90 % and 74.52 %, respectively. This approach can increase the accuracy of local tree density simulations, which is crucial for the precise modeling of the forest carbon sequestration potential, the development of targeted forest conservation strategies, and the implementation of effective carbon management practices. |
|---|---|
| ISSN: | 1470-160X |