Predicting Aboveground Carbon Storage in Different Types of Forests in South Subtropical Regions Using Machine Learning Models
ABSTRACT Motivated by the need to enhance the accuracy of forest aboveground carbon storage (ACS) assessments, this study aimed to explore the effectiveness of different machine learning models in predicting ACS across various subtropical forest types in southern China. The study was conducted in so...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2025-05-01
|
| Series: | Ecology and Evolution |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/ece3.71499 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849469775101558784 |
|---|---|
| author | Jiarun Liu Zihang Yang Lin Li Xiaoxue Chu Shiguang Wei Juyu Lian |
| author_facet | Jiarun Liu Zihang Yang Lin Li Xiaoxue Chu Shiguang Wei Juyu Lian |
| author_sort | Jiarun Liu |
| collection | DOAJ |
| description | ABSTRACT Motivated by the need to enhance the accuracy of forest aboveground carbon storage (ACS) assessments, this study aimed to explore the effectiveness of different machine learning models in predicting ACS across various subtropical forest types in southern China. The study was conducted in southern China, focusing on different types of subtropical forests. This region harbors several types of subtropical forests, which are rarely found at similar latitudes in the world. Variance inflation factor was employed to screen independent variables, resulting in the selection of 13 significant predictors. Four machine learning models—support vector machine (SVM), random forest (RF), multi‐layer perceptron (MLP), and extreme gradient boosting (XGB)—were constructed to estimate carbon storage. Model performance was evaluated using root mean square error, coefficient of determination (R2), and mean absolute error. The model with the best generalization ability was selected to calculate SHAP values for each predictor. The XGB model demonstrated superior performance across all forest types, with R2 values ranging from 0.898 to 0.974. In mountainous evergreen broad‐leaved forests, the prediction accuracy followed the order of XGB>MLP>SVM>RF. In valley rainforests, MLP showed the highest R2 value, but with higher MAE and RMSE, making it the second‐best choice. The RF model performed moderately, while the SVM model showed the poorest performance. The SHAP values indicated that maximum diameter at breast height, slope, mean DBH, species evenness, altitude, and maximum tree height had significant effects on ACS. XGB model exhibits the best prediction performance and strongest adaptability for estimating ACS in subtropical southern China forests. Additionally, the MLP model can serve as an effective model for assessing carbon storage in valley rainforests within this region. Machine learning methods provide valuable references for predicting and assessing ACS in different types of zonal forests. |
| format | Article |
| id | doaj-art-1b7133becf5a4b68933b403554875be2 |
| institution | Kabale University |
| issn | 2045-7758 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | Wiley |
| record_format | Article |
| series | Ecology and Evolution |
| spelling | doaj-art-1b7133becf5a4b68933b403554875be22025-08-20T03:25:20ZengWileyEcology and Evolution2045-77582025-05-01155n/an/a10.1002/ece3.71499Predicting Aboveground Carbon Storage in Different Types of Forests in South Subtropical Regions Using Machine Learning ModelsJiarun Liu0Zihang Yang1Lin Li2Xiaoxue Chu3Shiguang Wei4Juyu Lian5School of Life & Environmental Sciences Guilin University of Electronic Technology Guilin Guangxi ChinaSchool of Life & Environmental Sciences Guilin University of Electronic Technology Guilin Guangxi ChinaSchool of Life & Environmental Sciences Guilin University of Electronic Technology Guilin Guangxi ChinaSchool of Life & Environmental Sciences Guilin University of Electronic Technology Guilin Guangxi ChinaKey Laboratory of Ecology of Rare and Endangered Species and Environmental Protection, Ministry of Education – Guangxi Key Laboratory of Landscape Resources Conservation and Sustainable Utilization in Lijiang River Basin Guangxi Normal University Guilin Guangxi ChinaKey Laboratory of National Forestry and Grassland Administration on Plant Conservation and Utilization in Southern China, South China Botanical Garden Chinese Academy of Sciences Guangzhou ChinaABSTRACT Motivated by the need to enhance the accuracy of forest aboveground carbon storage (ACS) assessments, this study aimed to explore the effectiveness of different machine learning models in predicting ACS across various subtropical forest types in southern China. The study was conducted in southern China, focusing on different types of subtropical forests. This region harbors several types of subtropical forests, which are rarely found at similar latitudes in the world. Variance inflation factor was employed to screen independent variables, resulting in the selection of 13 significant predictors. Four machine learning models—support vector machine (SVM), random forest (RF), multi‐layer perceptron (MLP), and extreme gradient boosting (XGB)—were constructed to estimate carbon storage. Model performance was evaluated using root mean square error, coefficient of determination (R2), and mean absolute error. The model with the best generalization ability was selected to calculate SHAP values for each predictor. The XGB model demonstrated superior performance across all forest types, with R2 values ranging from 0.898 to 0.974. In mountainous evergreen broad‐leaved forests, the prediction accuracy followed the order of XGB>MLP>SVM>RF. In valley rainforests, MLP showed the highest R2 value, but with higher MAE and RMSE, making it the second‐best choice. The RF model performed moderately, while the SVM model showed the poorest performance. The SHAP values indicated that maximum diameter at breast height, slope, mean DBH, species evenness, altitude, and maximum tree height had significant effects on ACS. XGB model exhibits the best prediction performance and strongest adaptability for estimating ACS in subtropical southern China forests. Additionally, the MLP model can serve as an effective model for assessing carbon storage in valley rainforests within this region. Machine learning methods provide valuable references for predicting and assessing ACS in different types of zonal forests.https://doi.org/10.1002/ece3.71499aboveground carbon storage predictionfactor contributionsmachine learning modelsmulti‐layer perceptronrandom forestSHAP values |
| spellingShingle | Jiarun Liu Zihang Yang Lin Li Xiaoxue Chu Shiguang Wei Juyu Lian Predicting Aboveground Carbon Storage in Different Types of Forests in South Subtropical Regions Using Machine Learning Models Ecology and Evolution aboveground carbon storage prediction factor contributions machine learning models multi‐layer perceptron random forest SHAP values |
| title | Predicting Aboveground Carbon Storage in Different Types of Forests in South Subtropical Regions Using Machine Learning Models |
| title_full | Predicting Aboveground Carbon Storage in Different Types of Forests in South Subtropical Regions Using Machine Learning Models |
| title_fullStr | Predicting Aboveground Carbon Storage in Different Types of Forests in South Subtropical Regions Using Machine Learning Models |
| title_full_unstemmed | Predicting Aboveground Carbon Storage in Different Types of Forests in South Subtropical Regions Using Machine Learning Models |
| title_short | Predicting Aboveground Carbon Storage in Different Types of Forests in South Subtropical Regions Using Machine Learning Models |
| title_sort | predicting aboveground carbon storage in different types of forests in south subtropical regions using machine learning models |
| topic | aboveground carbon storage prediction factor contributions machine learning models multi‐layer perceptron random forest SHAP values |
| url | https://doi.org/10.1002/ece3.71499 |
| work_keys_str_mv | AT jiarunliu predictingabovegroundcarbonstorageindifferenttypesofforestsinsouthsubtropicalregionsusingmachinelearningmodels AT zihangyang predictingabovegroundcarbonstorageindifferenttypesofforestsinsouthsubtropicalregionsusingmachinelearningmodels AT linli predictingabovegroundcarbonstorageindifferenttypesofforestsinsouthsubtropicalregionsusingmachinelearningmodels AT xiaoxuechu predictingabovegroundcarbonstorageindifferenttypesofforestsinsouthsubtropicalregionsusingmachinelearningmodels AT shiguangwei predictingabovegroundcarbonstorageindifferenttypesofforestsinsouthsubtropicalregionsusingmachinelearningmodels AT juyulian predictingabovegroundcarbonstorageindifferenttypesofforestsinsouthsubtropicalregionsusingmachinelearningmodels |