Analyzing Risks in Voluntary Forest Carbon Offsets Using Open Data: A Hybrid Framework Integrating Retrieval-Augmented Generation in LLMs and Geospatial Analytics

Abstract

The credibility of voluntary carbon markets hinges on the quality of carbon offset projects, particularly in forestry and land-use sectors where claims of additionality and emissions reductions are often disputed. This paper introduces a novel, open-source approach to evaluating carbon offset projects by integrating open datasets, satellite-based remote sensing, and large language models (LLMs). Focusing on additionality and baseline integrity, the study examines existing challenges—including inflated baselines, inconsistent standards, leakage risks, and limited transparency—and proposes a system to automate early-stage project assessment. The platform combines AI-driven document analysis and geospatial data processing to evaluate risk factors such as additionality, leakage, and policy compliance, offering stakeholders an accessible, scalable tool to identify high-integrity carbon credits and mitigate greenwashing. This work aims to enhance transparency, accountability, and trust in the voluntary carbon market through data-driven, user-friendly decision support.