What does the tool do?
The tool computes the carbon footprint associated with round
trip flights, according to the methodology of several publicly
available calculators.
It does so for a set of trips from a given city of origin
to a set of destinations. Similarly, the tool allows to compute
the carbon footprint of a larger set of trips, corresponding to a
conference, a meeting and so on. For this, the city of departure
for each participant to the event has first to be provided.
If multiple destination cities are provided, the tool ranks those
cities according to the associated carbon footprint.
While online calculators enable to compute the footprint of a
limited number of trips, this tool enables to compute the
footprint of a larger number of trips in an automated way.
Furthermore, it provides an estimate based on data from seven
different methods, whose estimates can differ significantly. If
more than one method is selected by the user, the tool returns
the mean of the estimates of all selected methods.
How does the tool work?
A round trip is defined by a city pair. The two cities are
geolocated and from their longitude and latitude, the great
circle distance is computed. This is the shortest path a plane
can follow. Some methods thus consider uplift correction
factors to account for deviations from the shortest paths, e.g.,
when a plane avoids bad weather conditions. In
addition, two cities may not be connected by a direct flight.
This is accounted for by increasing by 5% the great circle
distance. Each method provides the carbon dioxide emission
in kg as a function of the flight distance in km. Thus from the
increased great circle distance, the carbon dioxide emission of
a flight associated with a trip between a city pair is computed
and multiplied by two to account for a round trip.
Which methods are used?
The tool incorporates seven different methods, among the most
widely used, and for which the methodology used is
documented (see additional resources later). This tool is by no means
a critical review of the different methods. It takes the methods as
they are described and presented. The methods may differ in their
assumptions (e.g. with or without radiative forcing included)
or in the perimeter considered (e.g. adding to the flight emissions,
the one associated with the production phase and transport of the kerosene).
The tool attempts to implement
each method to the highest possible level of accuracy.
Providing more than one method enables to get
a mean value, while illustrating the significant differences in
their estimates. In alphabetic order, the data considered are
from ADEME: French Agency for Ecological Transition, atmosfair: a German carbon offsetting company,
DEFRA: the UK Department for Environment, Food & Rural
Affairs, ICAO: International Civil Aviation Organization, from the
KLM carbon compensation service data and finally from
myclimate, a Carbon offseting company, used in particular by Lufthansa.
This list is obviously not exhaustive but represents
a variety of estimates from lower to higher values.
Updating the emission factors
The emission factors will be regularly updated. In April 2021, the emission factors
from ADEME were updated (a minor revision took place on september 2023, from the impact of CO2 web site).
The average emission factors for short, medium, long flights
are considered as asverages for planes of ~200 seats or more (for ADEME,
larger emissions for lower capacity planes). Very similar data are considered by the
Labos1.5 organisation. In September 2023,
the emission factors from DEFRA were updated. Similarly to the ADEME factors,
short-hauls have a distance less than 1000 km, medium-hauls between 1000 and 3500 km,
and long-hauls above 3500 km. The DEFRA emission factor for medius-hauls is the mean
of the factors for domestic flights and short-haul international flights.
The DEFRA emission factor for long-hauls is the mean between long-haul flights (to/from UK)
and international flights(to/from non-UK). The DEFRA 2023 data and ADEME 2021
data are consistent with one another within uncertainties (from 10% to 70% for
short and long flight emissions respectively).
How are the different methods built?
What is needed for each method is a function giving the
carbon dioxide emission as a function of the flight distance.
ADEME and DEFRA provide mean emission factors, as a
function of flight distance. Myclimate provides an analytical
formula. For ICAO and atmosfair, the on-line calculators have
been run for a wide range of flights of varying distances (~100
flights spanning from 300 km to 12000 km) and the estimates
have been fitted with linear functions, covering adjacent
distance intervals (the calculators were run in 2020).
For its carbon compensation service, KLM
provides on its web site a table of emissions for a large range
of flights. The KLM data have been also been fitted with linear
functions. Differences up to a few tens of percent may be found between
the data and the linearly interpolated values.
Accounting for non CO2 effects?
Aviation contributes to climate change more than just with the emission
of Carbon dioxide from burning fuels, by releasing gases and particles
directly into the upper troposphere and lower stratosphere where
they have an impact on atmospheric composition.
To quantify to total climate impact of burning fuels, the direct
CO2 emission of aircrafts is then multiplied by a factor, which
in the literature refers either to a Radiative Forcing Index (RFI)
or a Global Warming Potential (GWP) integrated over some time period
(a 100-year time horizon was adopted for the Kyoto Protocol to the
United Nations Framework Convention on Climate Change). There is a debate
in the science community on how non-CO2 effects should be modelled.
For instance, a RFI multiplier of 1.7 (update 2023) is used by DEFRA,
and a multiplier of 2 is considered by myclimate and ADEME
(see discussion in Jungbluth, N. & Meili,
C. Int J Life Cycle Assess (2019) 24: 404.).
ATMOSFAIR considers a multiplier of 3, for all emissions above 9 km,
accounting for the profile of the flight.
ICAO, KLM and the French Agency for Ecological Transition provides emission factors
with and without the impacts associated with the trails.
Therefore, the methods based on ICAO and KLM data are not recommended,
as they do not account for non-CO2 effects, but are still
given as methods providing the lowest emissions.
Seating category
The tool assumes economy seats in computing the travel footprint.
On average it can be considered that the
footprint can be multiplied by ∼ 1.5, ∼ 2.0 and ∼ 2.5
for flying in Premium Economy, Business and First class.
Accounting for train emission
The minimum distance for flying (one leg of the round trip)
is an input to be selected by the user (it is set to 500 km by default).
Below the minimum distance for flying, it is assumed that train is used.
Deviations from the shortest path is accounted by a
1.35 multiplication factor. The tool thus assumes the
mean of the emission factors of national and international rails,
as provided by DEFRA (i.e. 23 grams per passenger km). This value is consistent with the ADEME values,
if one considers trains powered by non-nuclear energy (it would be a factor of ~10 lower in that case).
To relate the duration of a train journey to a travel distance,
an average speed of 100 km/h is assumed.
Input and output data
The inputs are provided in US English for the city and country
names, without diacritics. On each line, the city and country
names must be separated by a comma. Pasting a CSV file in the
form is possible, provided that a comma separates
the city and country names.
A round trip is defined by a city pair. If the user enters cityA as
the origin city, and twice cityB as destinations, the tool returns
the cumulative emission and distance from two round trips
involving cityA and cityB, and indicates that 2 round trips were involved.
The same happens if the user enters twice cityA as the city of origin
and cityB as the sole destination.
Three types of inputs can be considered
depending on whether the user wants an “individual” estimate
or an estimate for a conference, meeting and so on. In the
former case, the “origin” city is unique and the “destination”
cities may be multiple. The tool returns the carbon dioxide
equivalent emission for each city of destination. The tool can thus
be run also for an organization interested to know about its travel
footprint, in which case the "origin" city is the city from which
employees travel. In the later case, the “origin” cities are multiple
(i.e. the cities from which the participants to the conference
depart from), and the “destination” city may be a single host city
or multiple host cities if the user wants to compare their
associated footprint. If there is one destination, the tool returns
the carbon dioxide equivalent for each city of origin. If multiple
destinations are provided the tool returns the carbon dioxide
equivalent emission summed over all cities of
origin and for each city of destination. The cumulative round trip
distance to each city of destination is also provided.
The result page provides a summary plot which can be
downloaded, as well as a csv and raw yaml file, which can be used for
further processing. The csv file lists the name of the
city as in the form, the address to which it
was geolocated, the carbon dioxide emission (in kg), the
distance travelled, the number of trips possible by train (i.e.
when the distance is less than the minimum flying distance,
e.g. 500 km) and the number of trips by plane. The plot and the
csv file rank the cities against the carbon dioxide emissions.
Trouble shooting
The estimation can go wrong if a city is not properly
geolocated. This may happen because the name of the city is
wrongly spelled or the geolocator (OpenStreetMap) is confused. An error
should be listed at the end of the result page. Don’t be
surprised, if the name recovered by the geolocator is not
exactly the one you had expected (e.g. a city is located at the
address of an embassy). An error may also occur if the input file
submitted does not comply with the requested format,
including font encoding. If nothing happens during a simple request, it
is most likely caused by the geolocator being unavailable. In
this case, try again a few minutes later.
In case of very large many-to-many origin-to-destination
combinations, the computing time may become very large. The advice is
to split the inputs into parts. This may happen when comparing tens
of thousands origins with hundreds of destinations.
If the format of the input data is not respected, the calculator will
crash, without being too informative. Please check your inputs first.
If you experienced an error, it will always help to empty the Cache of
your navigator. Alternatively, you may want to try with a different
browser.
Caveats
The numbers provided by the tool do not come with
uncertainties, and shall only be considered indicative of the true emission.
However, the numbers can be
used for relative comparisons, e.g. when comparing two cities
for hosting a conference.
Confidentiality
The data provided in the form will remain confidential, as will be the results.
Disclaimer
This tool is provided on a best effort basis as a service to
members of the science community. The numbers provided
are informative and have obviously no legal value.
Reference
Results from the tool may reference to Barret (2020),
Estimating, monitoring and minimizing the travel footprint
associated with the development of the Athena X-ray Integral Field Unit,
An on-line travel footprint calculator released to the science community,
Experimental Astronomy, 49, 183. doi:10.1007/s10686-020-09659-8.
Original motivation
Global warming poses a threat for life on our planet. Emissions of
carbon dioxide by aircrafts keeps increasing, as the world
economy keeps growing (it is about 3% of the anthropogenic
emissions nowadays). Carrying scientific research is generally associated
with traveling across the world, and thus air travel is likely to
dominate the carbon footprint of most scientists and is
likely to be large for developing international projects. This tool was
first developed to compute the travel footprint
associated with the development of the X-ray Integral
Field Unit to fly on board the Athena space observatory
in the mid 2030s. The rather large number derived,
typically one hundred ton of carbon dioxide per bi-annual consortium meeting,
imposed concrete actions to reduce the footprint
of the project, by re-considering the number of
large meetings, implementing different ways of interacting and
working collectively in a world-wide consortium. The tool was
further improved to easily compute the travel footprint associated
with individual traveling or with the organization of events, involving
a large number of travels. Finally, by comparing different,
widely used methods, providing so different estimates, the tool is
expected to raise awareness within the scientific community
(and hopefully the authorities and medias)
about the lack of regulations or framework on the critical
matter of estimating aircraft emissions.
Concluding note
As a personal note, I would like to stress that, as a scientist, I
find it very worrying that there are no standards
for computing the flight emissions. May this tool help to raise
awareness on this issue.
Additional resources