CarbonX: An Open-Source Tool for Computational Decarbonization Using Time Series Foundation Models

2025-10-01

AI Summary

Existing carbon intensity forecasting tools rely on grid-specific data and models, lacking global generalizability and uncertainty quantification—limiting their utility in carbon-aware computing. This paper introduces the first open-source decarbonization computing framework built upon a time-series foundation model (TSFM), enabling zero-shot cross-grid carbon intensity forecasting and imputation. Trained globally and equipped with calibrated uncertainty quantification, the model delivers high-accuracy forecasts up to 21 days ahead: zero-shot MAPE is 15.82% across 214 grids, and 9.59% (16.54% in tail regions) across 13 benchmark grids, with 95% prediction interval coverage. Imputation performance improves by 1.2–3.9× over baselines. Our core contribution is a unified, scalable, and confidence-aware global carbon intensity modeling paradigm.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Planning, Routing, and Scheduling: Planning with Language Models
• Computer Vision: Diffusion Models for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Sustainability and carbon-aware systems for Web, mobile, and WoT
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

Computational decarbonization aims to reduce carbon emissions in computing and societal systems such as data centers, transportation, and built environments. This requires accurate, fine-grained carbon intensity forecasts, yet existing tools have several key limitations: (i) they require grid-specific electricity mix data, restricting use where such information is unavailable; (ii) they depend on separate grid-specific models that make it challenging to provide global coverage; and (iii) they provide forecasts without uncertainty estimates, limiting reliability for downstream carbon-aware applications. In this paper, we present CarbonX, an open-source tool that leverages Time Series Foundation Models (TSFMs) for a range of decarbonization tasks. CarbonX utilizes the versatility of TSFMs to provide strong performance across multiple tasks, such as carbon intensity forecasting and imputation, and across diverse grids. Using only historical carbon intensity data and a single general model, our tool achieves a zero-shot forecasting Mean Absolute Percentage Error (MAPE) of 15.82% across 214 grids worldwide. Across 13 benchmark grids, CarbonX performance is comparable with the current state-of-the-art, with an average MAPE of 9.59% and tail forecasting MAPE of 16.54%, while also providing prediction intervals with 95% coverage. CarbonX can provide forecasts for up to 21 days with minimal accuracy degradation. Further, when fully fine-tuned, CarbonX outperforms the statistical baselines by 1.2--3.9X on the imputation task. Overall, these results demonstrate that CarbonX can be used easily on any grid with limited data and still deliver strong performance, making it a practical tool for global-scale decarbonization.

Problem

• Forecasting carbon intensity without grid-specific electricity data
• Providing global coverage using a single general model
• Generating uncertainty estimates for reliable carbon-aware applications

Innovation

• Uses Time Series Foundation Models for decarbonization tasks
• Provides zero-shot carbon intensity forecasts globally
• Delivers forecasts with uncertainty estimates and imputation