Ask three carbon calculators what a flight from Stockholm to New York emits and you can get three answers that differ by more than half. None of them is necessarily wrong. They answer different questions.
A distance-based calculator answers: what does an average flight of this length emit? A flight-based calculator answers: what did this particular flight emit? The gap between the two comes down to five things.
1. Which aircraft actually flew
Fuel burn per kilometre differs a lot between aircraft types and engine generations. The European Environment Agency publishes fuel consumption per aircraft type for the landing/take-off cycle and for the climb/cruise/descent phase, and that is the data our algorithm uses. A route served by both an older and a newer type can have two different answers depending on the day.
2. How many seats were in it
The same aircraft type is flown with different cabin layouts. A Boeing 737-800 is delivered with 189 seats in an all-economy layout and about 162 in a two-class layout.¹ The fuel burned is almost the same, so the emissions per seat differ by roughly 17 %. For wide-bodies the difference between airlines can be considerably larger.
3. How full it was
Passengers are a small share of an aircraft’s weight, so emissions per passenger fall almost in proportion to how full the flight is. Annual reports put some low-cost carriers at 84–93 % seat occupancy while other airlines sit around 76–83 %.¹ That alone is a difference of up to about 20 % per passenger.
4. Where it actually flew
Flights do not follow the great-circle line. No-fly zones, conflict areas, weather, airspace congestion and approach patterns all add distance. On most routes the detour is modest; on some it more than doubles the distance. A calculator that starts from the straight line cannot see this.
5. Whether it cruised above 8 000 metres
Contrails and other non-CO₂ effects only form at altitude and in cold, ice-supersaturated air. A turboprop on a short hop never gets there; a long-haul jet spends most of the flight there. Our algorithm applies a height factor only to the part of the flight above roughly 8 000 metres, so a short regional flight carries no uplift while a long-haul flight approaches ×1.9. Whether a method applies this factor at all is, by itself, a difference of up to 90 %. We explain the reasoning in The height factor explained.
How common methods handle these five
| Distance | Aircraft | Seats | Occupancy | Height factor | |
|---|---|---|---|---|---|
| ICAO Carbon Emissions Calculator | Great-circle distance plus a fixed correction by distance band | Average of the types scheduled on the route | Not considered separately | Average load factor from ICAO traffic data | Not included (CO₂ only) |
| UK DEFRA/DESNZ conversion factors | Three distance bands, +8 % for indirect routing | Not considered | Cabin-class multipliers | Built into the band average | Optional ×1.7 uplift |
| Carbon Compute | Distance actually flown | The aircraft that flew | The operating airline’s cabin layout, weighted by seat size | The operating airline’s occupancy | Applied to the part of the flight above ~8 000 m |
The first two are perfectly reasonable methods for an estimate. They are also deliberately insensitive to the things a traveller or a travel manager can influence: which airline, which aircraft, which cabin, direct or via a hub. That is why two trips on the same route get the same number.
What this means for your report
If you report business travel under GHG Protocol scope 3 category 6, you have to pick a method and be able to defend it. Three practical consequences:
- Don’t mix methods between years. A switch from distance bands to per-flight data can move the total by tens of percent without a single trip changing.
- Keep the flight-level data. Flight number, date and cabin class let an auditor reproduce every figure.
- State the height factor. Report CO₂e with the factor you use and say what it is; many organisations also disclose the CO₂-only figure alongside.
¹ Schennings, A., Larsson, J. & Robèrt, M. (2019). Development and implementation of an emission optimization model for passenger flight bookings. Sustainable Environment Research 29:25. doi:10.1186/s42834-019-0024-5
Method descriptions are based on the ICAO Carbon Emissions Calculator Methodology (v13, 2024) and the UK Government greenhouse gas conversion factors (DESNZ/DEFRA).