Carbon Sink Variations and Driving Factors Based on XGBoost-SHAP Analysis in the Beijing-Tianjin-Hebei Region
XGBoost-SHAP分析に基づく北京-天津-河北地域における炭素吸収源の変動と駆動要因 (AI 翻訳)
Yiquan Liu
🤖 gxceed AI 要約
日本語
北京・天津・河北（京津翼）地域の炭素吸収源の動態をXGBoostとSHAPで分析。GPPとNPPは有意に増加し、北高南低の空間分布を示した。主要な制限要因としてGPPには相対湿度、NPPには気温が影響。地域の炭素吸収源管理に科学的根拠を提供する。
English
This study analyzes carbon sink variations in the Beijing-Tianjin-Hebei region using XGBoost and SHAP. GPP and NPP show significant increasing trends with north-high south-low spatial heterogeneity. Relative humidity primarily limits GPP, while temperature limits NPP. Findings provide scientific evidence for regional carbon sink management and climate policy.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
日本の炭素吸収源管理にも応用可能なXGBoost-SHAP分析手法を提示。ただし対象地域は中国であり、日本の生態系への直接適用には調整が必要。
In the global GX context
This study provides empirical evidence on carbon sink trends in a key Chinese industrial region, relevant for global carbon cycle modeling and climate mitigation policy. The XGBoost-SHAP approach offers a robust method for identifying limiting factors.
👥 読者別の含意
🔬研究者:Provides a rigorous ML-based approach to carbon sink analysis that can be replicated for other regions.
🏢実務担当者:Land managers and forestry companies can use the identified driving factors to optimize carbon sequestration.
🏛政策担当者:Offers evidence for targeting humidity and soil moisture management in carbon sink policies.
📄 Abstract（原文）
The Beijing-Tianjin-Hebei (named Jing-Jin-Ji in China, JJJ), as a key economic and industry development region in China, plays a vital role in achieving China's carbon neutrality goals. Understanding the dynamic changes in JJJ's carbon sink and its driving factors is crucial for policy formulation on carbon emissions. In this study, we use multi source satellite observations, eXtreme gradient boosting model (XGBoost) and SHapley Additive exPlanations (SHAP) to explore the spatiotemporal variations of gross primary productivity (GPP) and net primary productivity (NPP), and the impact of environmental factors on them. The results show that both GPP and NPP exhibited significant upward trends (p < 0.001) in BTH, with increasing rates of 9.79 g C m-2 year-1 and 5.21 g C m-2 year-1, respectively. The spatial distribution of GPP and NPP exhibits significant similar north-high, south-low heterogeneity. High-value are concentrated in the forest and mixed forest ecosystems, while low-value are primarily located in farmlands and urban construction. High-performance XGBoost models (R2 equals 0.72 for GPP and 0.84 for NPP) were established and employed to implement SHAP analysis. We found that the primary limiting environmental factor affecting GPP is relative humidity, followed by third-layer soil moisture and temperature. Temperature is the primary limiting driver for NPP, followed by temperature-radiation interactions and third-layer soil moisture. This study provides scientific evidence for understanding the carbon cycle processes within ecosystems in the JJJ region and their response to climate change, offering significant reference value for regional carbon sink management.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.54254/2755-2721/2026.mh35286first seen 2026-07-17 05:13:52
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。