ナイジェリアにおける分解型エネルギー供給に基づくエネルギー関連CO2排出量の予測
Predicting energy-related carbon dioxide emissions based on disaggregated energy supply in Nigeria (原題)
Abubakar Jamilu Baita
🤖 gxceed AI 要約
日本語
ナイジェリアのエネルギー供給を化石燃料・再生可能エネルギー等に分解し、機械学習(ラッソ・リッジ回帰)でエネルギー関連CO2排出量を予測した研究。テスト期間ではリッジ回帰がRMSE・MAPE・MSE最小、R2最大で最良の予測性能を示した。石油・ガスが排出量に最も影響し、2024〜2030年に再生可能エネルギーが増えてもCO2排出は約10%増加すると予測している。
English
This study predicts Nigeria's energy-related CO2 emissions using disaggregated energy supply factors and machine learning. Ridge regression outperformed lasso on test data (lowest RMSE/MAPE/MSE, highest R2), with petroleum and gas driving emissions most. Despite renewable growth, emissions are projected to rise ~10% from 2024 to 2030, prompting policy recommendations.
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
新興国ナイジェリアの排出予測・エネルギー転換シナリオは、日本の企業がサプライチェーン排出(Scope 3)や新興国市場の脱炭素リスクを評価する際の参考になる。ただし日本の政策・開示制度との直接の接点は乏しい。
In the global GX context
Offers an emerging-market emissions forecasting case relevant to global transition-finance and country-level decarbonization pathway analysis, though it does not engage TCFD/ISSB disclosure frameworks directly. Useful as comparative evidence on how fossil-fuel-dependent economies project emissions under renewable targets.
👥 読者別の含意
🔬研究者:機械学習による国別排出予測の手法比較(ラッソvsリッジ)と変数重要度の実例として参考になる。
🏢実務担当者:新興国サプライヤー・市場の排出トレンドをScope 3や移行リスク評価に組み込む際の示唆を得られる。
🏛政策担当者:再生可能エネルギー目標だけでは排出増を抑えられない可能性を示し、化石燃料依存国での追加政策の必要性を裏付ける。
📄 Abstract(原文)
Adequate supply of renewable and nonrenewable energies drive economic growth and promote energy security. While renewable energies drives green energy transition, supply of fossil fuels cause high environmental pollution, which increase climate change. To mitigate the harmful effects of fossil fuels and improve sustainability, Nigeria set renewable energy targets to increase their penetration in the overall energy mix. Therefore, this study aims to predict energy-related CO2 emissions in Nigeria by incorporating disaggregated energy supply factors. The findings reveal that lasso linear regression model performs very well during the training period. However, it exhibits overfitting feature during the testing period. In contrast, ridge regression has better predictive performance based on the testing dataset. It has the lowest RMSE, MAPE, and MSE, as well as the highest R2. As well, the variable importance evaluation shows that PCI and gas have the largest impact on energy-related CO2 emissions. By building energy transition scenario, this paper forecast future CO2 emissions trends from 2024 to 2030. Despite projected growth of renewable energy, CO2 emissions is expected to rise by about 10 percent during the forecast period. The author proffers policy recommendations for curtailing energy emissions in Nigeria.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.68171/j.cofeco.2026.23first seen 2026-11-08 04:46:41 · last seen 2026-11-13 04:52:44
🔔 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。