A carbon figure is the easiest number on an aviation site to get wrong, and the hardest for a reader to check. Type a route into any of a dozen calculators and you will get a dozen answers, sometimes differing by a factor of three, all presented with the same confident styling. The difference is almost never arithmetic. It is the assumptions buried underneath: which aircraft, how full, whose seat, and whether the number counts carbon dioxide alone or tries to price in the rest of aviation's warming.
So rather than print a figure and ask you to trust it, this page sets out the entire chain we use, in order, including the places where the honest answer is a range and the one step that rests on an assumption rather than a measurement. If you disagree with a step, you will be able to see exactly which one and adjust the result yourself.
The chain, end to end
The calculation behind every flight page is deliberately short. More steps would mean more places to hide a guess.
- Take the great-circle distance between the origin and destination airports.
- Look up the measured fuel burn, in kilograms per kilometre, for the aircraft type that normally flies the route.
- Multiply the two to get the fuel burned.
- Multiply the fuel by 3.16 to get carbon dioxide.
- Divide across the passengers actually carried, weighting each seat by the floor area it occupies.
Steps one through four are measurement and arithmetic. Step five is where the judgement lives, and we come back to it below.
The difference between carbon calculators is almost never arithmetic. It is the assumptions buried underneath.
Why fuel burn is read, not modelled
The single biggest lever on the result is how much fuel the aircraft actually burns, and this is where a lot of calculators quietly invent something. It is tempting to model it: take a maximum takeoff weight, apply a drag coefficient, integrate over a cruise profile. The output looks rigorous and is often wildly off, because real fuel burn depends on engine variant, winglets, cruise altitude, airline weight policy and the age of the airframe.
We do not model it. For each aircraft type we hold published measurements of fuel burn at specific sector lengths, taken from operator and manufacturer reporting, and we look the route up against them. A Boeing 787-9 has published figures at several distances; a route lands between two of them and we interpolate.
Burn per kilometre is not a constant, which is the reason the sector length matters at all. A longer flight has to carry the fuel that burns the fuel, so a 12,000 km sector burns more per kilometre than a 6,000 km one in the same airframe. A very short sector is inefficient for the opposite reason: proportionally more of it is spent climbing, which is the most fuel-hungry part of any flight. Applying one average figure to every route would flatter long flights and penalize short ones.
When a route falls outside the measured range
Sometimes a route is shorter or longer than any published measurement for that type. In that case we hold the nearest measurement flat rather than extrapolating past it, and the flight page says so in plain words. Running a straight line beyond the last real data point would manufacture precision exactly where the aircraft is least typical. Where we do this, the page also tells you which direction the true figure is likely to sit, so a held-flat number is never presented as though it were interpolated.
A few types carry no figure at all. If the reference tables do not measure a variant, we do not substitute a near-enough relative: an A340-600 is not an A340-300, and borrowing the latter's numbers would produce something that looks exactly as authoritative as a sourced figure while being nothing of the kind. Freighters get no per-passenger number either, for the obvious reason.
Did you know?
The 3.16 figure is not an industry estimate or an average of operator reports. It falls straight out of chemistry: burning a kilogram of kerosene combines its carbon with atmospheric oxygen, and the resulting carbon dioxide weighs a little over three times the fuel did, because the oxygen comes from the air. ICAO, the UK government conversion factors and the IPCC inventory guidelines all use the same number for the same reason.
Dividing one total across a cabin
The aircraft burns the same fuel whoever is on board. A cabin does not each carry its own emissions; it divides one fixed total. The question is how.
The approach used by the UK government conversion factors, and by us, is floor area. A seat's share of the flight is the share of the cabin it takes up, so a flat bed that occupies the space of roughly three economy seats carries roughly three economy seats' worth of the fuel. The weights work out at about 2.9 times economy for business and 4 times for first.
| Cabin | Floor-area weight | Share of one flight's CO2 |
|---|---|---|
| Economy | 1.0 | Below the cabin average |
| Premium economy | 1.6 | Slightly above economy |
| Business | 2.9 | Roughly three economy seats |
| First | 4.0 | Roughly four economy seats |
This is why the economy figure on a flight page sits below the per-passenger average printed just above it. The average divides the total by every passenger equally; the cabin split recognizes that the passenger in seat 2A has taken several times as much of the aircraft. Both numbers are correct, and they answer different questions.
The one assumption in the chain
To split a cabin you need to know how many seats are in each class, and that is the one input we do not hold as data. Airlines configure the same aircraft type very differently: a long-haul carrier's Airbus A380 and a holiday charter's are the same airframe with wildly different interiors.
So the split we use is the layout typical of that size of aircraft, and every flight page prints the seat counts it assumed rather than hiding them. If it says 410 economy, 89 business and 26 first and you know your airline flies a denser cabin than that, you can see immediately what to discount. A carrier flying all-economy would put every passenger close to the economy figure.
How full the aircraft is
Emissions per passenger depend on how many passengers there are. We use a load factor of 83 percent, close to the global average the airline industry reports, rather than assuming a full aircraft. A full aircraft would make every airline look better than it is, since the fuel is burned whether or not the seats are sold.
We also use the typical multi-class seat count rather than the maximum certified capacity. Those two numbers can be far apart. An A380 is certified to carry up to 853 people in an all-economy layout that no airline has ever operated, while real three-class examples hold around 525. Dividing a fixed fuel burn by imaginary seats is one of the quieter ways to halve a carbon figure.
What the figure deliberately leaves out
The number on a flight page counts carbon dioxide. It is not a full climate impact figure, and we would rather say so than let a precise-looking number imply completeness.
- Non-CO2 effects. Contrails and nitrogen oxides emitted at altitude also warm the climate. The European Commission assessment of aviation non-CO2 effects puts their combined contribution at roughly the same size again as the carbon dioxide alone. Multipliers for this are genuinely contested, so we publish the part that is settled and tell you what is missing.
- Belly freight. We charge every kilogram to passengers. A widebody on a long route also carries cargo that would fairly take a share, so on those routes our per-passenger number sits at the cautious end.
- The day's actual routing. We use great-circle distance. Real flights fly airways, take weather deviations and hold. Real burn moves with winds and cruise altitude too.
- Ground operations. Taxi, auxiliary power and the airport's own footprint are not counted.
If you want to see how much the real world departs from a clean great-circle line, the live flight map is the fastest illustration: tracks bend around weather and airspace constantly. Our guide to what contrails actually are covers the largest of the non-CO2 effects in more depth.
Key takeaways
- Fuel burn is read from published measurements per aircraft type and sector length, never modelled from first principles.
- Carbon dioxide is fuel multiplied by 3.16, a figure fixed by chemistry rather than by industry estimate.
- Cabins divide one fixed total, weighted by the floor area each seat occupies.
- The class seat split is the only assumption in the chain, and every page prints the seat counts it used.
- Figures use an 83 percent load factor and typical multi-class seating, not maximum certified capacity.
- The number counts CO2 only; contrails and NOx roughly double the real climate effect.
Frequently asked questions
Why is your number different from my airline's calculator?
Usually because of the cabin split, the load factor or the non-CO2 question. Airlines often quote an economy figure at a high load factor, which is the most flattering combination available. Compare like with like: check whether their number is per passenger or per seat, which cabin it assumes, and whether it counts carbon dioxide only.
Does flying business class really emit three times as much?
Your presence on the aircraft does not burn three times the fuel, but your share of it is roughly three times larger, because you occupy roughly three economy seats' worth of cabin. The fuel is fixed; the question is how it is fairly divided, and floor area is the standard answer.
What does the tree figure mean?
It is the number of trees that would take that much carbon dioxide back out of the air over a year, using the US Environmental Protection Agency figure of 0.06 tonnes absorbed per medium-growth urban tree annually. It describes absorption, not destruction. A flight does not consume trees.
Why do some flight pages have no CO2 figure?
Either we have no published fuel burn for that aircraft variant, or we have not yet observed which aircraft operates that flight, or the aircraft is a freighter with no passenger cabin. We leave the section out entirely rather than substituting a similar airframe.
Can I see the figure for a specific flight?
Yes. Every tracked flight page carries the calculation for that route and aircraft. Start from the flight number directory or search for a callsign, and the emissions section sits below the route details.
None of this makes a carbon figure exact, and it is worth being honest that no method available produces an exact one. What it can do is be legible: a reader who wants to know why a number looks the way it does should be able to follow it all the way down to the measurement it rests on. That is the standard we are trying to hold to here, and the reason this page exists at all.
