A question came up over a beer with friends last night about whether traveling by plane or auto would create less carbon dioxide per mile.
Giles (2005) reports that a Boeing 767 traveling from London to New York emits about 1 tonne (103 kg) per passenger. London is about 3,600 miles from New York, so this comes out to about 0.28 kg/mile = 0.61 lb/mile.
One gallon of gasoline weighs about 8 lb and is about 80% carbon, so a gallon contains about 6.4 lb of carbon. When oxidized to form CO2, 6.4 lb of carbon creates 6.4 x (44/12) = 23.5 lb CO2/gallon. An auto that gets 25 miles/gallon uses 0.04 gallons/mile. Therefore the CO2 emitted by driving this auto is 0.04 gallons/mile x 23.5 lb CO2/gallon = 0.94 lb/mile.
Therefore riding in a (presumably full) Boeing 767 on a long flight emits about 2/3 as much CO2 per mile as driving (presumably alone) an auto having efficiency of 25 miles/gallon.
Sobering note to frequent fliers: If you are a member of the 100,000 mile/year club you contribute at least 30 tons (61,000 lb) of CO2 per year to global anthropogenic emissions. About 30% of this will still be in the atmosphere 200 years from now and 10% will remain in 400 years. So if you are a member of the 100,000 mile club, 3 tons of your 2005 emissions will still be in the atmosphere in the year 2405, all the while contributing a proportionate amount to greenhouse gas absorption of infrared radiation.
References
Giles, J., 2005: Europe set for tough debate on curbing aircraft emissions. Nature, 436, 764-765.