Tianyi Guan

About

This project examines how historical greenhouse-gas estimates change when the same public emissions database is updated. Using EDGAR v8.0, EDGAR 2024, and EDGAR 2025, the study compares CH₄ and N₂O estimates at the country or territory × exact IPCC 2006 sector × year level. The primary analysis covers 1990–2020, while 2021–2022 is treated separately as a recent, vintage-sensitive period. The study evaluates whether release-to-release revisions exceed ordinary adjacent-year variation within a fixed release and whether database version choice changes estimated trends, trend classifications, country rankings, or the apparent concentration of revision effects. The overall finding is a stable centre accompanied by a consequential tail: most comparable records change little, but some revisions materially affect downstream interpretations.

Keywords

EDGAR; greenhouse-gas inventories; methane; nitrous oxide; historical revisions; release sensitivity; database versioning; IPCC sectors; trend classification; country rankings; environmental informatics; reproducibility; data lineage;

Study Design

Cross-release comparisons are restricted to records that are structurally comparable across the three EDGAR versions. Raw blanks remain missing rather than being converted to zero; national totals are kept separate from exact-sector components; fossil and biogenic contributions retain source-cell lineage through aggregation; and only records with compatible structural partitions and numerical-contribution signatures enter the primary comparisons. A source-integrity issue affecting CH₄ IPCC 2006 sector 1.B.1 in EDGAR v8.0 is handled by symmetric isolation from cross-release primary metrics rather than by deletion, truncation, absolute-value transformation, or taxonomy substitution. No further details can be disclosed at this time.