What a flight's carbon number actually represents
An aircraft burns fuel; the calculation divides that fuel among the people on board. Kerosene releases about 3.16 kg of CO2 per kilogram burned, so if you knew the fuel uplift, the payload split and the passenger count you could compute a flight exactly. Nobody has those figures for a specific booking, so the standard approach works backwards from fleet statistics: total fuel burned by a class of aircraft on a class of route, divided by total passenger-kilometres flown, giving a factor in kilograms of CO2e per passenger-kilometre.
That factor is not constant with distance, and the shape of the curve surprises people. Short flights are worse per kilometre because take-off and climb consume a disproportionate share of the fuel and are amortised over very few cruise miles. Very long flights are also penalised, because the aircraft must carry the weight of its own fuel for hours before burning it. The efficiency sweet spot sits in the medium-haul range, which is why the factor used here falls from about 0.246 kg per passenger-kilometre on very short domestic hops to about 0.148 on long haul.
The second big lever is the floor space your seat occupies. Fuel burn is a property of the aircraft, and the per-passenger share follows the fraction of the cabin you take up. A lie-flat business seat occupies roughly three times the floor area of an economy seat, so it carries roughly three times the fuel. This is the reason a business-class booking on the same aeroplane, on the same day, carries several times the footprint of the economy seat one row back.
From two coordinates to a kilogram figure
Step one is the great-circle distance. The haversine formula computes the shortest path across a sphere from two latitude-longitude pairs. The half-angle form is used rather than the naive spherical law of cosines because it stays numerically stable for short distances, where the cosine version loses precision.
Step two is a routing uplift. No aircraft flies the great circle. It follows published airways, avoids restricted airspace, takes advantage of or fights jet-stream winds, and sometimes holds before landing. Established reporting methodologies add roughly 8 to 10 percent to the great-circle figure to represent this, and the calculator lets you set the percentage.
Step three applies the haul-band factor. The calculator selects the factor from the one-way great-circle distance: under 500 km, 500 to 3,700 km, and above 3,700 km. Notice that the band is chosen by the length of a single leg, not the round trip — a return journey is two flights of the same length, not one flight of double the length.
Step four applies the cabin multiplier, and step five, optionally, applies a radiative forcing multiplier. Aviation warms the climate through more than CO2: contrails that form persistent cirrus, nitrogen oxides that alter ozone and methane, and water vapour injected into the upper troposphere. The best current assessments put aviation's total effective radiative forcing at roughly twice its CO2-only forcing, and the UK Government's reporting factors offer an uplift of 1.9 for organisations that choose to include it. The uncertainty band around that figure is wide, and the non-CO2 effects are short-lived while the CO2 persists for centuries, so the two are not straightforwardly interchangeable. The calculator reports both numbers, and you should quote which one you mean.
Worked example: New York to London and back in economy
You fly JFK to Heathrow and back, in economy, once a year. JFK sits at 40.6413° N, 73.7781° W and Heathrow at 51.4700° N, 0.4543° W.
- Great-circle distance. Δφ = 10.8287° = 0.188997 rad and Δλ = 73.3238° = 1.279660 rad. The haversine term is sin²(0.094499) + cos(0.709325)·cos(0.898323)·sin²(0.639830) = 0.008903 + 0.472674 × 0.356479 = 0.177403. Taking 2 × 6,371.0088 × arcsin(√0.177403) gives 5,539.7 km one way.
- Haul band. 5,539.7 km is above 3,700, so the long-haul economy factor of 0.148 kg CO2e per passenger-km applies.
- Routing uplift and legs. 5,539.7 × 1.08 × 2 legs = 11,965.7 passenger-kilometres.
- Economy CO2e. 11,965.7 × 0.148 = 1,770.9 kg, or 1.77 tonnes for the round trip.
- With radiative forcing at 1.9. 1,770.9 × 1.9 = 3,364.7 kg, about 3.36 tonnes.
- In business class. Multiply the CO2e by 2.9: 1,770.9 × 2.9 = 5,135.6 kg without the RF uplift, or 9,757.8 kg with it.
- Offset cost. At $20 a tonne, offsetting the 3.36-tonne RF-inclusive economy figure costs $67.30.
For scale, the economy round trip without the RF uplift is roughly 40 percent of the annual emissions of driving 12,000 miles at 25 MPG, which the driving CO2 calculator puts at 4.27 tonnes. One business-class transatlantic return exceeds a whole year of that driving.
Typical round-trip footprints by route length and cabin
| One-way distance | Band factor (kg/pax-km) | Economy | Premium economy | Business | First |
|---|---|---|---|---|---|
| 300 km | 0.246 | 159.4 kg | 255.0 kg | 462.2 kg | 637.6 kg |
| 1,000 km | 0.156 | 337.0 kg | 539.1 kg | 977.2 kg | 1,347.8 kg |
| 2,500 km | 0.156 | 842.4 kg | 1,347.8 kg | 2,443.0 kg | 3,369.6 kg |
| 5,500 km | 0.148 | 1,758.2 kg | 2,813.2 kg | 5,098.9 kg | 7,033.0 kg |
| 9,000 km | 0.148 | 2,877.1 kg | 4,603.4 kg | 8,343.6 kg | 11,508.5 kg |
| 14,000 km | 0.148 | 4,475.5 kg | 7,160.8 kg | 12,979.0 kg | 17,902.1 kg |
Every cell is distance × 1.08 × 2 legs × band factor × cabin multiplier. The 300 km row shows the short-flight penalty clearly: it emits nearly half what a 1,000 km flight does over less than a third of the distance.
How to read the result
Say which number you are quoting. The two figures the calculator returns — with and without radiative forcing — differ by nearly a factor of two, and comparing an RF-inclusive figure from one calculator against a CO2-only figure from another is the single most common source of confusion in this subject. Corporate inventories under the GHG Protocol usually report CO2e without an RF uplift; consumer-facing calculators usually include one.
Compare against something concrete. A single long-haul return in economy is roughly one to two tonnes of CO2e. Set against a typical household footprint in the household carbon footprint calculator, two or three long-haul trips a year usually outweigh every other discretionary choice a household makes.
Treat cabin class as a real decision variable. Downgrading from business to economy on a long-haul return cuts your per-passenger figure by roughly two thirds — a bigger single change than most people can make anywhere else in their life in one booking.
Do not over-trust the precision. These are fleet-average factors. The actual emissions of your flight depend on aircraft type, load factor, winds, routing and freight carried in the hold, and a specific flight can plausibly differ from the average by a third in either direction. Quote the result to two significant figures and treat it as an estimate of magnitude.
Offsets are not a subtraction. Buying an offset does not reduce the emissions the flight caused; it funds an activity intended to reduce or remove an equivalent amount elsewhere, with real questions about additionality and permanence. Report gross emissions, then report offsets separately, which is what every serious reporting framework now requires.
Where flight estimates go wrong
- Using the round-trip distance to select the haul band. A 2,000 km return is two short-haul flights, not one 4,000 km long-haul flight, and the factors differ.
- Forgetting connections. A routing through a hub can add hundreds or thousands of kilometres over the direct great-circle path. Sum the legs individually rather than using origin-to-final-destination distance.
- Comparing an RF-inclusive number with a CO2-only number. Almost a factor of two, and completely invisible unless both sources state their scope.
- Assuming a full aircraft. Emission factors assume typical load factors. On a half-empty flight your true share is higher; you have no way of knowing, which is part of why these figures are estimates.
- Ignoring the freight in the hold. Passenger aircraft carry cargo, and allocation methods differ in how much of the fuel burn they assign to it. Different calculators make different choices here.
- Counting an offset as a reduction. Report gross emissions first. An offset is a separate financial transaction, not an adjustment to the physical number.
Where the emission factors come from
The distance-band factors used here are in the range published by national reporting programmes, notably the UK Government's annual greenhouse gas conversion factors for company reporting, which give separate figures for domestic, short-haul international and long-haul international flights by cabin class. Those factors are revised every year as fleet efficiency and load factors change, so if you are preparing a formal disclosure, take the current year's published values from the relevant national dataset rather than relying on the round numbers in a general-purpose calculator.
Alternatives, and when the comparison actually favours flying
On short routes, flying is usually the worst option per passenger. A 300 km flight emits roughly 80 kg per passenger one way; the same distance by rail is typically a small fraction of that, and a car with three or four occupants comes in well below the flight per person. The calculator's driving comparison uses a solo driver at 25 MPG over the great-circle distance, which is deliberately conservative in both directions — real roads are longer, but real cars often carry more than one person.
On long intercontinental routes there is usually no alternative at all, which changes the question from mode choice to trip choice: fewer, longer stays rather than frequent short ones. Two week-long trips emit twice what one two-week trip does, for the same days away.
Within an organisation, flights normally sit in Scope 3 category 6 (business travel) under the GHG Protocol, alongside employee commuting in category 7. They are reported separately from the Scope 1 fuel a company burns in its own vehicles and the Scope 2 electricity it buys — the same electricity that the electricity CO2 calculator quantifies. Keeping those categories separate matters more than the precision of any single flight estimate, because inventories are compared year over year and a category boundary that moves invalidates the trend.