This page is for those of you who want to know the details of where our numbers come from and what they mean. There are many topics. We've tried to order them from more significant to more esoteric.
Fuel Extraction, Production, and Transport Emissions not Included
Fuel is burned to extract, produce, and transport the fuel used in a journey. When we quote an emissions number we only consider the greenhouse gases released by the journey itself. We don't consider the additional emissions created prior to the use of the fuel in the journey. We made this choice to keep the site simple by making relative comparisons between different itineraries rather than trying to nail down an absolute number for a single itinerary.
This overhead factor is roughly the same for jet fuel as gasoline and so our relative comparisons of itinerary impacts remains accurate. It would be important to include these secondary emissions if we were comparing the impact of traveling to other impacts such as food consumption.
Since we aren't considering extraction, production, and transportation of fuel we use the EPA calculation which states that burning of a gallon of gas releases 8.89 kgs of CO2.
What exactly do the "Travel Modes" map to?
To keep the site simple we have chosen a couple of canonical types of travel. The table below summarizes what each of the travel modes means.
| Travel Mode | Description |
|---|---|
| Average Jet | Boeing 737 for ranges up to 4000 Nautical Miles Boeing 747 for ranges from 4000 to 7500 Nautical Miles |
| Modern Jet (efficient) | Boeing 737-Max for ranges up to 4000 Nautical Miles Boeing 787 for ranges from 4000 to 7500 Nautical Miles |
| Regional Jet (inefficient) | Embraer E170 Valid up to a range of 2000 Nautical Miles |
| Average Car | 25 mpg (eg Toyota Highlander) |
| Efficient Car | 40 mpg (eg Toyota Prius V) |
| Inefficient Car | 16 mpg (eg Toyota Sequoia) |
Why haven't you included electric cars?
We do plan on including electric cars in the future, but unfortunately their warming effects are difficult to include in our current relatively simple warming model.
Typically the warming impact of driving an electric vehicle is lower than driving a gas powered vehicle, but it is still quite substantial. The challenge of modeling electric vehicles is that a substantial portion of their lifetime warming impact occurs at the time of manufacture and less impact occurs during their operation. To create a valid equivalency between electric and gasoline powered vehicles we would have to model the warming impact of their manufacture. And then if we created a more sophisticated vehicle warming model that includes manufacture we would also then have to model the warming impact of the construction of aircraft. We are currently not up to that level of modeling complexity.
Radiative Forcing and FlyBetter's Philosophy of Exactness (or not).
Scientists are trying to come to an agreement on a model for radiative forcing due to flights. There seems to be broad agreement that the basic warming due to CO2 emissions should be increased as a way to account for the other warming effects of flight apart from the direct release of greenhouse gasses. The amount that basic emissions should be multiplied by is not yet nailed down.
We have choosen the radiative forcing number 3 which is at the high end of the range of current estimates. This means that if a flight is modeled to have released 100 kgs of CO2_e greenhouse gasses for a particular itinerary, the warming effect of that trip becomes 300 kgs of CO2_e once radiative forcing is added in.
This figure comes from the Stanford Scope 3 white paper. We chose a number at the higher end of the range because we aren't trying to make our model perfect. Instead in our tool we are trying to bracket the range of possibilities. Without radiative forcing our warming estimates for flying are almost certainly too low. If you include radiative forcing in our model, the resultant estimate is on the highest end of potential warming. The true impact of the flight will be somewhere in between.
It isn't critical for Google's travel impact model to add in radiative forcing because the relative comparisons between flight itineraries will remain similar, with and without radiative forcing. But it is important for our users to be able to include radiative forcing to be able to better compare flying to driving.
So what is your current model?
We believe Google's Travalyst impact model to be quite good in most aspects. We mainly follow this model using the following simplifications
- We limit the choice of aircraft to a couple of distinct types.
- We don't customize loading and load factor to the specific flight.
- Instead of using seating class factors from IATA RP 1726, we used United and Luftansa aircraft configurations from seat guru. (Listed below.)
- We don't have access to PianoX data and so we've had to estimate the efficency of the 737-Max as compared to the 737.
We added radiative forcing to our model to allow for better comparisons between flying and driving.