中国都市における化石燃料エネルギーとクリーンエネルギーの協調的移行発展水準の時空間的進化と駆動メカニズム：地理空間説明可能AIからの洞察
Spatio-temporal evolution and driving mechanism of the level of coordinated transitional development between fossil fuel energy and clean energy in Chinese Cities: Insights from geospatial explainable artificial intelligence (原題)
Yu Zhu, Lihong Zhao, Jin-liang Wang
🤖 gxceed AI 要約
日本語
中国291都市を対象に、2011〜2021年の化石燃料とクリーンエネルギーの協調移行度（CCD）を結合協調度モデルで測定。XGBoostとGeoShapleyを統合した地理空間説明可能AI（GeoXAI）で空間的異質性を分析した。CCDは全都市で上昇し、顕著な空間的自己相関を示し、原発立地アクセス性が全国・地域・都市の全スケールで最重要因子（寄与37.33%）と判明。多層的なエネルギー移行政策への示唆を提供する。
English
Using a coupling coordination degree model, this study measures coordinated fossil-fuel/clean-energy transition across 291 Chinese cities (2011–2021). A GeoXAI framework combining XGBoost and GeoShapley reveals spatial heterogeneity, with CCD rising everywhere and strong spatial autocorrelation. Nuclear plant accessibility is the dominant driver at all scales (37.33% nationally). Findings inform multi-scale energy transition policy and China's Dual Carbon goals.
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
中国の「双炭」目標と都市単位のエネルギー移行を扱うが、日本のSSBJ・有報・投資家対応への直接的接続は薄い。ただしAIによる空間的脱炭素評価手法は、日本の地域別GX政策や自治体の移行計画策定に応用可能な分析枠組みとして参考になる。
In the global GX context
Adds city-level, spatially explicit empirical evidence to the global energy-transition literature, complementing national-scale studies. The GeoXAI approach offers a transferable method for subnational decarbonization assessment relevant to ISSB/TCFD-aligned regional transition planning.
👥 読者別の含意
🔬研究者:XGBoostとGeoShapleyを組み合わせた空間的説明可能AIのエネルギー移行分析への応用事例として参考になる。
🏢実務担当者:地域別のエネルギー移行進捗とその駆動因子を把握するための空間分析手法として活用可能。
🏛政策担当者:都市・地域スケールでの協調的エネルギー移行政策設計と原発立地などインフラ配置の政策判断に示唆を与える。
📄 Abstract（原文）
Against the background of global warming, the coordinated transitional development between fossil fuel energy (FFE) and clean energy (CE) has become a core issue in China's energy transition. Most existing studies have analyzed energy transition at the national or provincial scale from a single dimension, overlooking city-level spatial heterogeneity and the need for an integrated FFE-CE perspective. Taking 291 Chinese cities as research objects, this study applies the coupling coordination degree (CCD) model to measure the level of coordinated transitional development during 2011–2021 and constructs a geospatial explainable artificial intelligence (GeoXAI) framework integrating XGBoost and the GeoShapley algorithm to analyze the spatially heterogeneous contributions of predictor variables to CCD predictions at national, regional, and city levels. The results show that: (1) CCD increased across all 291 cities during 2011–2021, with significant spatiotemporal differentiation. (2) CCD exhibited significant positive spatial autocorrelation (Global Moran's I > 0.6, Z > 2.58, p < 0.001), with High-High clusters concentrated in the eastern coastal region and Low-Low clusters emerging at the central-western border. (3) The XGBoost model achieved high prediction performance (test-set R 2 = 0.9481, MAE = 0.0142, RMSE = 0.0206), and nuclear power plant accessibility (NPPA) was the most important factor at all scales, contributing 37.33% of total GeoShapley importance nationally and up to 42.79% in the Eastern Region. (4) Factor interactions were markedly stronger in the Western Region than those at other scales, with the NPPA-Temperature interaction reaching 8.62 × 10 −4 . The findings of this study not only provide scientific guidance for the formulation of multi-scale energy coordinated transition policies, but also promote the implementation of the “Dual Carbon” goals in China.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.1016/j.jclepro.2026.149719first seen 2026-10-17 04:50:02
🔔 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。