GX & Decarbonization Research

Xu P.

This paper evaluates the causal effect of China's emissions trading scheme (ETS) using high-frequency trading data and double machine learning (DML). It estimates the impact of policy intervention on carbon prices and trading volumes, provi…

Jakkaraju V.T.D.

This paper proposes Predictive Green FinOps, a method for jointly optimizing cost, carbon, and reliability in AI-intensive cloud environments. It uses predictive analytics to dynamically adjust resource allocation, reducing carbon footprint…

Singh V.

This paper applies FinBERT, a financial domain-specific BERT model, to conduct sentiment analysis on ESG disclosures of Indian mutual funds using a pillar-based classification approach. The study demonstrates the effectiveness of NLP techni…

Khan T.

This paper proposes a conceptual framework for using AI in Scope 3 GHG emissions measurement, reporting, monitoring, and assurance from a critical systems perspective. It examines AI's potential to improve data quality and automation while …

Feiya Luo, Yanyun Nian, Pinqi Rao +2

This study proposes a scalable workflow integrating UAV LiDAR and hyperspectral data for individual tree species classification and carbon stock estimation in urban forests. Using an adaptive feature selection method (ACV-DCC) and RF classi…

Jacob Goldverg, Elvis Rodrigues, Tevfik Kosar

This paper proposes spatiotemporal scheduling methods for data transfers to minimize carbon footprint, leveraging AI/optimization techniques and accounting for temporal and geographical variations in carbon intensity. It contributes to ener…

Ivan Savin (5189054), Mateo López Carel, Eva Schlindwein

This paper uses regression to link topic prevalences extracted from ESG reports to environmental KPIs. It demonstrates how NLP and machine learning can quantitatively assess the relationship between disclosure content and actual environment…

Ivan Savin (5189054), Mateo López Carel, Eva Schlindwein

Using computational linguistics on 1,477 ESG reports from STOXX Europe 600, this study identifies 34 main topics (six environmental). It finds that topics like sustainable value chains and renewable energy correlate with improved environmen…

Laurence L. Delina, Johanne Rei R. Castro, Yuet Sang Marie Tung

This chapter explores the use of AI, including a generative AI chatbot, to facilitate just energy transitions for vulnerable communities. It proposes strategies grounded in energy justice, reliable datasets, and participatory decision-makin…

Xavier Martínez-Barbero, Ana Pastor-Merino, Josep Domenech

This paper presents a nationwide dataset of 104,553 Spanish firms, using LLMs to extract energy transition practices (efficiency, decarbonization, renewables) from corporate websites. Aggregated at province, sector, and size levels for 2023…

Mengxi Xie

This dissertation combines machine learning (NLP) with econometrics to empirically analyze environmental policies. The first three studies examine the EU's Carbon Border Adjustment Mechanism (CBAM) via meta-analysis, US equity market reacti…

Susanna Ruth Gruyal

This paper critically analyzes the IMF's proposed International Carbon Price Floor (ICPF) on AI carbon emissions through a law and political economy lens. It argues that market-based efficiency assumptions mask inequities, disadvantaging sm…

Ahmed G. Abo-Khalil

This paper proposes a unified framework integrating AI-driven renewable forecasting, grid economics, and lifecycle assessment. Using deep learning, it reduces prediction errors by 50% and operational costs by 18.7%. The study includes AI en…

Das N.

This paper proposes an end-to-end framework integrating IoT, AI, and Layer-2 blockchain for real-time MRV (Monitoring, Reporting, Verification) and autonomous tokenization of carbon credits in industrial CCUS. AI is employed for automated v…

Agraj Magotra, Md. Rafiqul Islam Rana, F. S. Shishir +1

This paper proposes a machine learning and NLP pipeline for analyzing ESG and sustainability disclosures in the textile and apparel industry, applicable to supply chain disclosure analysis.

Thi-Kien Dao, Trong-The Nguyen

This paper proposes CASO, a framework for carbon-aware VM placement integrating adaptive RBF surrogate model with self-adaptive PSO-DE swarm optimizer. It minimizes carbon emissions, energy, SLA violations, and latency simultaneously under …

Kálita Cristina Araújo, Ana Carolina Fernandes Maciel, Bruno B. F. da Costa

This systematic review (PRISMA-based) examines whether automating CO2 emission calculation with AI in BIM can support low-carbon building design. From 2567 records (2021-2025), 85 studies were classified as Core (BIM+CO2+AI) or Base. 60% qu…

Chris Meniw

This whitepaper analyzes the dual role of AI: optimizing power grids and integrating renewables while increasing energy and carbon footprint from compute infrastructure. It examines deployments in smart grids, industrial optimization, and r…

Lee H, Jung J, Roh J

This study proposes a Residual U-Net deep learning model to correct significant wave height forecasts from WAVEWATCH III, validated with real-ship fuel consumption data. The corrected forecasts show improved accuracy up to 7-8 days ahead, e…

Hanno Blankenstein

This paper presents a practical approach combining AI-driven visual analytics with automated drone-based inspection workflows for scalable emissions reduction and energy transformation in energy production. It overcomes limitations of tradi…