European emissions and the energy transition: a data guide

If you want to measure Europe's energy transition, the raw material is all public — Eurostat, ENTSO-E, the European Environment Agency. The problem is that "how green is country X" is not one number. It is a dozen numbers, published by different bodies, on different grains, in different units, each with a footnote that changes what it means. Here is the map, and where the bodies are buried.

Greenhouse-gas emissions: the headline and its footnotes

The number everyone quotes is total greenhouse-gas emissions, expressed in CO2-equivalent so that methane and nitrous oxide can be added to carbon dioxide on one scale. Eurostat publishes it per country per year, split by economic sector.

Three footnotes decide whether your comparison is honest:

• LULUCF. Land use, land-use change and forestry can be a *sink* (negative emissions). Whether it is included flips national totals by whole percentage points. "Emissions excluding LULUCF" and "including LULUCF" are different series; never mix them in one chart.
• Base year. Reduction targets are quoted against 1990. If your dataset starts in 2008 you cannot compute the headline "% below 1990" at all.
• Territorial vs consumption. Official inventories are *territorial* — they count emissions produced inside the border, not those embodied in imports. A country that offshored its heavy industry looks greener than its consumption is.

Our European GHG Emissions dataset keeps the sector split and the LULUCF distinction explicit, so you are never silently comparing two different definitions.

"Renewable share" is at least three different numbers

Under the Renewable Energy Directive methodology (Eurostat's SHARES), the renewable *share* is reported four ways, and they diverge hard:

• Overall share of gross final energy consumption — the target headline.
• Electricity — usually the highest, because wind and solar land here first.
• Transport — usually the lowest, dragged down by liquid fuels.
• Heating and cooling — the quiet one that decides whether a country actually decarbonises, because heat is most of final energy.

A country can post an impressive electricity number and a dismal heating number. If a dashboard shows one "renewable %" without saying which, it is hiding the part that matters.

Carbon intensity of power: the real-time transition signal

Annual shares tell you the trajectory. Grams of CO2 per kWh of electricity tells you what is happening this hour. You compute it by weighting the generation mix — from European generation by fuel — by per-technology emission factors. It is the metric that moves when the wind drops and a country fires up gas or coal to keep the lights on, and it is the one an emissions-aware scheduler or a green-tariff product actually consumes.

Energy import dependency: the security dimension

The transition is not only about carbon; it is about who you buy fuel from. Net energy import dependency — imports minus exports as a share of gross available energy — is the security counterpart, and after 2022 it became a front-page number. Our European Energy Dependency dataset carries it in total and by fuel (solid fuels, oil, natural gas), so you can see exactly which import a country is exposed to.

The circular economy: the other half of "green"

Decarbonisation gets the headlines; material circularity is the quieter half. Three indicators do most of the work: municipal waste generated per capita, the municipal-waste recycling rate, and the circular material-use rate — the share of material fed back into the economy rather than extracted new. They sit in our European Circular Economy dataset and join to the rest on country and year.

The join problem — and how we solve it once

Every series above is ultimately "country by year," but they arrive on incompatible grains: GHG inventories are annual with a two-year lag, SHARES is annual, ENTSO-E generation is sub-hourly, import dependency is annual. Units clash (ktCO2e, %, GWh, gCO2/kWh). Codes clash (ISO-2 here, TSO bidding zones there). Stitching them into one comparable panel — one row per country per year, UTC-anchored where time matters, units documented — is a week of careful, quietly error-prone work.

That panel is exactly our European Energy Transition dataset: renewable share, carbon intensity, power and gas price, GHG and import-dependency, rolled to country and joined on one key. You analyse the transition; you do not spend the first week reconciling definitions. Every series names its Eurostat / ENTSO-E upstream and its licence, so it is safe to publish, resell, and train on.