About Google Flight’s New CO2 Display

Google Flights now displays a carbon emission (CO2) estimate for most flights.

Folks, this is a game changer because it brings flight-related emissions data to the shopping table on a scale that reaches millions of flight shoppers.

Google Flight display for Seattle to London, one way on Nov.1st 2021

Not only is Google Flight’s CO2 data credible and easy to understand, it is easy to see which flight is better or worse, and by how much. This has big implications, namely:

  • Leisure and unmanaged business travelers will become accustomed to seeing this data. Many will use it to select their flights, making CO2 data part of the “consumer experience”.
  • Business travelers who shop first on Google Flights before booking in the OBT/TMC channel will get used to seeing this data. Many will use it to select their flights.
  • Travel managers will be expected by travelers and their executives to provide booking tools that provide flight-specific CO2 data.
    • This will create a standards problem, unless the booking tools use the same CO2 model as does Google Flights. More on this below.
  • Airlines will feel pressure to improve their CO2 metrics as they appear in Google Flights, especially if Google is successful at making its CO2 model a free resource to others, as it says it plans to do.
  • Unfortunately, it will reinforce the belief that premium seats, e.g., business class, should be allocated more CO2 per passenger than economy seats. This belief is wrong, and leads to more flight emissions, not less.

Insights About Google Flight’s CO2 model

Google’s CO2 model calculations are based on the European Environmental Agency (EEA) framework 3.1a, published in 2019, and some of its related data sources, including the BADA flight emissions model, which may use 2016 aircraft and engine data. The EEA framework is described here. Google apparently uses the Tier 3A approach, described on page 25. Google also uses data from other sources, including those from airlines and other sources to obtain the aircraft models and seating configurations.

Some things that caught my eye about Google’s approach:

More connections don’t necessarily mean more emissions. Note in the exhibit above that a 2-stop flight (circled in blue) has less emissions than the one-stop flight below it.

There seems to be no adjustment made for a flight’s passenger load factor or cargo weight, nor for the actual in-flight routing or actual gate-to-gate time. Factoring in these variables would make the model more accurate, IF the model was given each flight’s actual data. This data is practically impossible to get pre-flight. Other flight CO2 models may include these variables by making good-faith estimates, but they are still estimates, and probably a good example of providing false precision. Google’s approach is good enough.

That said, Google uses a traditional method for allocating more CO2 to premium economy, business and first class seats, based on the square footage of these types of seats. I’ll take another tilt at the windmill by saying again this is fundamentally wrong, and harmful to the climate.

Moving on. A flight’s emissions are shown as better or worse based on the other flights’ emissions for the same O&D, the same route, and the same cabin, and the same day(s) of travel. If a flight’s emissions are within 5%, plus or minus, of the median emission value, it is marked as “average”. Users can sort the itineraries based on CO2, so it makes it easy to pick the least-emitting flight.

Google says it plans to make its model available to the travel industry for free via the Travelyst coalition. It would be even better if Google published an API tomorrow to let anyone consume their calculations. The quicker the travel industry coalesces around flight and hotel CO2 calculation methods, the better. Until then, expect travelers to be confused about seeing significantly different emission estimates for the same flight. Confusion leads quickly to distrust, which is not what we want on this issue.

Innovation Opportunities

A few quick thoughts on how this platform could be put to even better use:

  • Change the basis of allocating CO2 from the seat’s size, to the amount of CO2 emitted for the weight of the seat, its passenger and the passenger’s luggage, per square foot.
  • Publish an API so anyone could consume these calculations (as mentioned above).
  • Integrate this CO2 data into your corporate booking tool.
  • Use the API and your TMC’s historical booking data for your program to set a 2019 baseline of flight emissions. Better, have the bright folks at Thrust Carbon or eco.mio do the heavy lifting, and enable ways to make offsetting easy and ethical.
  • Update the analytical modeling to use more current data on engine emissions.
  • For those concerned about the accuracy of the CO2 estimates, lobby for the airlines to publish for every flight the
    • Amount of fuel consumed;
    • Cargo weight;
    • Passenger and baggage weight;
    • Passenger load factor (wishful thinking).

Flight emissions will only grow in importance. It’s good to see Google enabling better decisions on this front.

1 thought on “About Google Flight’s New CO2 Display

  1. Scott – great points all around. I agree with you on the need for shared standards, weight based modeling, and especially for the call for people to lobby for more usable and consistent data points. To solve this global issue, we need engagement on all fronts, with companies of all sizes – and to your point, to get that level of engagement we need standards that build trust through transparency and accuracy.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.