Open-source traffic and CO2 emission dataset for commercial aviation

This record is a global open-source passenger air traffic dataset primarily dedicated to the research community. It gives a seating capacity available on each origin-destination route for a given year, 2019, and the associated aircraft and airline when this information is available. 
Context on the original work is given in the related article (https://journals.open.tudelft.nl/joas/article/download/7201/5683) and on the associated GitHub page (https://github.com/AeroMAPS/AeroSCOPE/).A simple data exploration interface will be available at www.aeromaps.eu/aeroscope.The dataset was created by aggregating various available open-source databases with limited geographical coverage. It was then completed using a route database created by parsing Wikipedia and Wikidata, on which the traffic volume was estimated using a machine learning algorithm (XGBoost) trained using traffic and socio-economical data. 
1- DISCLAIMER
The dataset was gathered to allow highly aggregated analyses of the air traffic, at the continental or country levels. At the route level, the accuracy is limited as mentioned in the associated article and improper usage could lead to erroneous analyses. 
Although all sources used are open to everyone, the Eurocontrol database is only freely available to academic researchers. It is used in this dataset in a very aggregated way and under several levels of abstraction. As a result, it is not distributed in its original format as specified in the contract of use.
As a general rule, we decline any responsibility for any use that is contrary to the terms and conditions of the various sources that are used. In case of commercial use of the database, please contact us in advance.
2- DESCRIPTION
Each data entry represents an (Origin-Destination-Operator-Aircraft type) tuple.
Please refer to the support article for more details (see above).
The dataset contains the following columns:

"First column" : index
airline_iata : IATA code of the operator in nominal cases. An ICAO -> IATA code conversion was performed for some sources, and the ICAO code was kept if no match was found.
acft_icao : ICAO code of the aircraft type
acft_class : Aircraft class identifier, own classification.

WB: Wide Body
NB: Narrow Body
RJ: Regional Jet
PJ: Private Jet
TP: Turbo Propeller
PP: Piston Propeller
HE: Helicopter
OTHER
seymour_proxy: Aircraft code for Seymour Surrogate (https://doi.org/10.1016/j.trd.2020.102528), own classification to derive proxy aircraft when nominal aircraft type unavailable in the aircraft performance model.
source: Original data source for the record, before compilation and enrichment.

ANAC: Brasilian Civil Aviation Authorities
AUS Stats: Australian Civil Aviation Authorities
BTS: US Bureau of Transportation Statistics T100
Estimation: Own model, estimation on Wikipedia-parsed route database
Eurocontrol: Aggregation and enrichment of R&D database
OpenSky
World Bank
seats: Number of seats available for the data entry, AFTER airport residual scaling
n_flights: Number of flights of the data entry, when available
iata_departure, iata_arrival : IATA code of the origin and destination airports. Some BTS inhouse identifiers could remain but it is marginal.
departure_lon, departure_lat, arrival_lon, arrival_lat : Origin and destination coordinates, could be NaN if the IATA identifier is erroneous
departure_country, arrival_country: Origin and destination country ISO2 code. WARNING: disable NA (Namibia) as default NaN at import
departure_continent, arrival_continent: Origin and destination continent code. WARNING: disable NA (North America) as default NaN at import
seats_no_est_scaling: Number of seats available for the data entry, BEFORE airport residual scaling
distance_km: Flight distance (km)
ask: Available Seat Kilometres
rpk: Revenue Passenger Kilometres (simple calculation from ASK using IATA average load factor)
fuel_burn_seymour: Fuel burn per flight (kg) when seymour proxy available
fuel_burn: Total fuel burn of the data entry (kg)
co2: Total CO2 emissions of the data entry (kg)
domestic: Domestic/international boolean (Domestic=1, International=0)

3- Citation
Please cite the support paper instead of the dataset itself. 

Salgas, A., Sun, J., Delbecq, S., Planès, T., & Lafforgue, G. (2023). Compilation of an open-source traffic and CO2 emissions dataset for commercial aviation. Journal of Open Aviation Science. https://doi.org/10.59490/joas.2023.7201