Cao et al. (2025) Influencing Factor Analysis of Vegetation Spatio-Temporal Variability in the Beijing–Tianjin–Hebei Region Based on Interpretable Machine Learning

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Forests
• Year: 2025
• Date: 2025-12-18
• Authors: Yuan Cao, Lanxuan Guo, Hefeng Wang, Anbing Zhang
• DOI: 10.3390/f16121873

Research Groups

Not explicitly stated in the provided text.

Short Summary

This study integrated multi-source data and machine learning methods to simulate and analyze Normalized Difference Vegetation Index (NDVI) changes in the Beijing–Tianjin–Hebei (BTH) region over the past two decades, identifying climate and human activities as key drivers with varying spatio-temporal importance across land use types.

Objective

• To quantitatively understand vegetation driving mechanisms across spatio-temporal scales by simulating and analyzing Normalized Difference Vegetation Index (NDVI) changes in the Beijing–Tianjin–Hebei (BTH) region, attributing the relative importance of climate and human activities across different land use types.

Study Configuration

• Spatial Scale: Beijing–Tianjin–Hebei (BTH) region.
• Temporal Scale: Two decades (2000 to 2020).

Methodology and Data

• Models used: XGBoost (eXtreme Gradient Boosting) for prediction, SHapley Additive exPlanations (SHAP) for predictor importance quantification. Compared against Random Forest (RF), Support Vector Machines (SVM), and K-Nearest Neighbors (KNN).
• Data sources: Multi-source data including Normalized Difference Vegetation Index (NDVI), climate variables (precipitation, temperature), and human activity indicators.

Main Results

• The XGBoost model demonstrated excellent performance in simulating NDVI changes from 2000 to 2020, achieving a coefficient of determination (R²) > 0.96, Mean Absolute Error (MAE) < 0.02, and Root Mean Square Error (RMSE) < 0.027, outperforming RF, SVM, and KNN.
• Vegetation in the BTH region showed an overall improving trend, with 47.96% of the total area exhibiting significant improvement.
• Precipitation, temperature, and human activities were identified as the most significant predictors of NDVI, with their relative importance varying over time and exhibiting clear spatial heterogeneity in NDVI responses.
• Primary predictors differed by land use type: NDVI in cropland and grassland was mainly driven by precipitation, forest NDVI by temperature, and urban/built-up areas by human activities.

Contributions

• Developed a novel analytical framework that integrates nonlinearity and spatial heterogeneity for a quantitative "overall-categorical" analysis of important predictors behind NDVI changes.
• Provided a new methodological reference for attributing vegetation dynamics.
• The findings offer valuable insights for implementing classified regulation and promoting ecological restoration in the BTH region.

Funding

Not explicitly stated in the provided text.

Citation

@article{Cao2025Influencing,
  author = {Cao, Yuan and Guo, Lanxuan and Wang, Hefeng and Zhang, Anbing},
  title = {Influencing Factor Analysis of Vegetation Spatio-Temporal Variability in the Beijing–Tianjin–Hebei Region Based on Interpretable Machine Learning},
  journal = {Forests},
  year = {2025},
  doi = {10.3390/f16121873},
  url = {https://doi.org/10.3390/f16121873}
}