Global aviation emissions have been growing despite international efforts to limit climate change. Quantifying the status quo of domestic and international aviation emissions is necessary for establishing an understanding of current emissions and their mitigation. Yet, a majority of the United Nations framework convention on climate change (UNFCCC)-ratifying parties have infrequently disclosed aviation emissions within the international framework, if at all. Here, we present a set of national aviation emission and fuel burn inventories for these 197 individual parties, as calculated by the high-resolution aviation transport emissions assessment model (AviTeam) model. In addition to CO2 emissions, the AviTeam model calculates pollutant emissions, including NOx, SOx, unburnt hydrocarbons, black carbon, and organic carbon. Emission inventories are created in aggregated and gridded format and rely on Automatic Dependent Surveillance–Broadcast combined with schedule data. The cumulative global fuel burn is estimated at 291 Tg for the year 2019. This corresponds to CO2 emissions of 920 Tg, with 306 Tg originating from domestic aviation. We present emissions from 151 countries that have yet to report their emissions for 2019, which sum to 417 TgCO2. The improved availability of national emissions data facilitated by this inventory could support mitigation efforts in developed and developing countries and shows that such tools could bolster sector reporting to the UNFCCC.
1.Introduction
With climate change affecting societies worldwide, global efforts to curb anthropogenic emissions are being taken (Shukla et al 2022, SPM, p 14, Lee et al 2023, p 5). The United Nations framework convention on climate change (UNFCCC) established an international treaty in 1992 as the first of a series of international agreements by the United Nations targeting climate change (United Nations 1992). The UNFCCC reporting requirements contain principles on inventory creation, and a distinction between industrialised countries (Annex I countries), which are required to submit inventories on greenhouse gases annually, and developing countries (non-Annex I countries), which have fewer reporting obligations (United Nations 1992). As of today, the UNFCCC has 198 signing parties, whereof one is the European Union and 197 parties that are called countries in the context of this paper (UNFCCC 2023b). Within the UNFCCC reporting, aviation CO2 emissions are regularly reported by Annex I countries and voluntary by non-Annex I countries and separated into domestic and international emissions. The subsequent Kyoto Protocol from 1997 is an international agreement that sets binding emission reduction targets based on values reported under the UNFCCC (UNFCCC 1997). Following up, the 2015 Paris Agreement aims goal to limit global warming to below 2 ∘C with a 1.5 ∘C warming compared to preindustrial temperatures as the target (UNFCCC 2015). The primary means to achieving this target are emission reduction pledges compared to 1990 by participating nations called National determined contributions (NDCs). In these agreements, emissions are allocated to the country of emission (territory-based allocation).
Despite national pledges and international agreements, CO2 emissions from aviation have been increasing, and aviation is estimated to have contributed 2.4% to global anthropogenic CO2 emissions in recent years (Lee et al 2021). Non-CO2 emissions from aviation fuel combustion comprise nitrogen oxides (NOx ), sulfur oxides (SOx ), carbon monoxide (CO), non-volatile particulate matter (nvPM), and organic carbon (OC). These emissions also affect the climate, increasing aviation's contribution to human-induced global warming to date to about 4% (Klöwer et al 2021). Non-CO2 emissions can also affect air pollution and human health (Eastham and Barrett 2016, World Health Organization 2021), the environment (Sand et al 2016), and contribute to condensation trail formation (Lee et al 2021).
The international nature of aviation poses challenges to legislators and mitigation efforts in addition to being widely recognised as a 'hard-to-abate' sector for technological reasons (Jaramillo et al 2022, Mayer and Ding 2022). One challenge is that territory-based allocation is not straightforward to apply where international flights cross borders and overfly third countries. A second challenge is that air travel demand has a distribution highly skewed towards affluent countries and individuals in the top 1%-percentile of emitters, causing as much as 50% of global aviation emissions (Gössling and Humpe 2020). Hence, demand and territory-based allocation may give largely different results. Both challenges may have motivated the exclusion of international aviation and shipping from the Kyoto Protocol (Lyle 2018), which affects an estimated 70% (the share of international flights' fuel burn) of all aviation emissions (Teoh et al 2023).
The International Civil Aviation organization (ICAO) serves as a United Nations entity responsible for international coordination in the sector. However, its existing climate strategy has drawn critique for perceived insufficiency in ambition and means to meet the Paris Agreement's goal (Lee 2018, Lyle 2018, Pathak et al 2022, Fuglestvedt et al 2023, p 99). Hence, national efforts formulated in the NDCs could play an important role in mitigating aviation emissions. Yet, we note that many parties' NDCs currently do not explicitly encompass either domestic or international aviation (https://unfccc.int/NDCREG), while other parties, such as the European Union, the United Kingdom, Switzerland and New Zealand, include both domestic and international aviation emissions in their general mitigation planning (Mayer and Ding 2022). Countries committed to quantifiable mitigation targets for aviation emissions have an incentive to maintain regular aviation emission inventories (similar to the UNFCCC Annex I countries) to track their progress.
Aviation emission modelling building on individual flight trajectories in combination with flight physics-based fuel burn models is an established practice to create emission inventories (IPCC 2006, Wilkerson et al 2010, Wasiuk et al 2016). The introduction of Automatic Dependent Surveillance–Broadcast (ADS-B) technology has opened a new, high-resolution data source in this field. ADS-B technology is a surveillance technology that entails that aircraft periodically broadcast their position and other flight parameters. The usability of ADS-B data for emission inventory creation has been demonstrated in several studies (Klenner et al 2022, Quadros et al 2022, Zhang et al 2022, Teoh et al 2023), some using only the origin-destination information in the ADS-B data (Quadros et al 2022). Some recent studies provide global emission inventories (Quadros et al 2022, Teoh et al 2023), while a few national inventories exist (van Pham et al 2010, Klenner et al 2022, Zeydan and Şekertekin 2022, Zhang et al 2022, Puliafito 2023). However, there is no recent inventory that uses ADS-B locations for emission locations and provides a consistent set of national inventories for all countries compatible with UNFCCC standards.
In this work, we address the need to establish robust quantifications of aviation's contribution to individual countries' emissions, as this could promote a further inclusion of aviation into NDCs. We introduce a set of national emission inventories with global coverage that concurrently functions as a set of national inventories aligned with UNFCCC standards (Tier 3B methodology from IPCC 2006). In section 2, we describe how we use the aviation transport emissions assessment model (AviTeam) established in Klenner et al (2022) to create inventories of domestic and international aviation emissions for all 197 UNFCCC-ratifying countries. We describe a fuel burn and emissions inventory based on ADS-B telemetry data and how we extract aggregated emissions inventories for economically developing (non-Annex I) and Annex I countries. In section 3, we inspect the generated UNFCCC-compliant data that uncover the contribution of each country's domestic and international aviation emissions (allocated to the airport of departure). We conduct a benchmarking exercise with other aviation emission inventories, demonstrating how both overall emission levels and allocation of emissions to nations agree with existing inventories. Section 4 discusses the role of the study in informing decision-makers engaged in reducing national emissions.
2.Methods
We calculate aviation fuel burn and emissions with the (AviTeam). In the following sections 2.1–2.4, we describe the AviTeam model, the global flight data, the process of flight trajectory creation, and, eventually, the conversion from flight trajectories to emission inventories.
2.1.The AviTeam model
The model was introduced by Klenner et al (2022), and posterior improvements are described in Klenner et al (2023). Here, we provide a short summary and updates for the global version. The AviTeam computes fuel burn for individual flights for the entire flight envelope based on a given or simulated geospatial trajectory, usually provided by ADS-B data, using the base of aircraft data 3 version 15 (BADA 3) model (Nuic et al 2010). The BADA 3 model uses simplified flight physics in combination with a parametrisation of more than 250 aircraft types. Each aircraft type is assigned the representative engine used in BADA 3. Combining fuel burn, atmospheric conditions assuming the standard atmosphere, and engine information allows estimating emissions of the species CO2, H2O, SOx , OC (all linear to fuel burn), NOx , HC, CO, and nvPM (all non-linear to fuel burn). NOx , HC, and CO emissions are calculated using the boeing fuel flow method 2 (Dubois and Paynter 2006). In the context of aviation emission modelling, non-volatile particulate matter (nvPM) mass and BC mass are often understood as synonyms (Abrahamson et al 2016, Quadros et al 2022). nvPM mass and number emission indices (EI nvPMm , EI nvPMn ) are estimated using the latest available nvPM method (T4/T2 method), described by Teoh et al (2023). If relevant nvPM measurements are not given for a specific engine, EI nvPMm values are calculated with a method that correlates nvPM emissions with the smoke number reported in the ICAO engine emission database (Peck et al 2013, Federal Aviation Administration 2021, ICAO 2023a). If neither approach is applicable, an EI nvPMm of 30 mg kg-fuel−1 is used (Federal Aviation Administration 2021, Quadros et al 2022). For EI nvPMn , if the T4/T2 is not applicable, a default value of 
particles kg-fuel−1 is used (Teoh et al 2023).
Compared to the previous AviTeam model version in Klenner et al (2022), a new algorithm to estimate aircraft masses is implemented. Take-off-mass (
) is now explicitly estimated based on operating empty weight (
), payload (
) and mission fuel (
):
In addition, the aircraft mass is updated iteratively in each timestep by subtracting the previous fuel burn. Further information on the model update is presented in the supplementary material of Klenner et al (2023).
2.2.Global flight data
Global flight movement data are sourced from ADSBexchange (adsbexchange.com). The global ADS-B telemetry data set for 2019 contains trajectory information, aircraft identifiers, and callsigns. As the spatial coverage of ADS-B data varies (Teoh et al 2023, SI, p 4), we complement ADS-B data with schedule data from cirium aviation analytics (www.cirium.com/), which is described below. The final data set contains 36.2 million individual flights for 2019. Flights contain information on departure and arrival airports, scheduled departure time, estimated arrival time, aircraft type, aircraft (hex code), airline and trajectory. Flights with several stops are divided into individual legs.
To align the airport information with the UNFCCC countries, we identify the departure airports' countries with the help of the airport database of ourairports.com. The European Union is represented by the individual member states. Disputed areas like Western Sahara are included in international departing traffic, but no national datasets are extracted, and neither are flights allocated to other countries. We define flights and their emissions as domestic if the departure and arrival airports are in the same country. This sometimes entails domestic aviation emissions occurring above foreign territory (compare Klenner et al 2022). Information on seat configurations per aircraft type is provided in the supplementary section C.10.
2.3.Flight trajectory creation
The combination of ADS-B and scheduled flight information follows these steps:
1.
ADS-B raw data processing with division into individual flights
2.
Matching of ADS-B trajectories and scheduled flight data
3.
Completion of matched ADS-B trajectories with fuel burn calculation
4.
Filtering of flights with the removal of conflicting flights
5.
Repeat matching of processed ADS-B trajectories and scheduled flight data
In Step 1, raw ADS-B data available in a 60 s resolution are grouped by aircraft identifier, and separated into individual flight legs using a combination of callsign information and altitude and speed profiles. Further, missing data points along the trajectory are added along the shortest great-circle distance. In Step 2, ADS-B trajectory data are combined with schedule information. Missing ADS-B trajectories are simulated using a surrogate flight that was operated on the same route and by the same aircraft, if possible. If no surrogate flights are found, flights are simulated using idealised flight trajectories. These flight trajectories are simulated using idealised climb and descent profiles contained in BADA 3 and cruise flights at a cruise altitude uniformly distributed around the cruise altitude suggested by BADA 3. In Step 4, flights that present irregular fuel consumption or emission indices are eliminated from the dataset. This filtering warrants 'Step 5: Repeat matching of processed ADS-B trajectories and scheduled flight data' to replace the eliminated flights. We refrain from computationally expensive weather matching, as excluding wind data is expected to introduce only a small error at a global scale of approximately 1% (Quadros et al 2022).
2.4.From flight trajectories to global aviation inventories
When applying the AviTeam to the global flight trajectory data, gridded emission inventories are created. Figure 1(a), shows the resulting global annual aggregated CO2 emissions. Given that flights' departure airports are contained in the global flight data, we can further disaggregate global emissions into different emission layers, as exemplified for the emissions in the area of Brazil and its surroundings (figures 1(b)–(e)). Total global emissions (figure 1(b)) are divided into emissions from international departures (figure 1(c)), domestic departures (figure 1(d)), and emissions caused by overflights and international arrivals (figure 1(e)). This approach is applied to all flights and all countries, providing 394 distinct emission layers (one domestic and one international departure layer for 197 countries) for each emission species.
The gridded data have a longitudinal and latitudinal resolution of 0.5∘, fifteen altitude levels, and are provided in annual aggregates. Higher resolutions can be made available as the inputted ADS-B data are available at a much higher resolution. A view of the global distribution of aviation CO2 emissions shows dense traffic in certain areas and flight corridors (figure 1). The latitudinal distribution reveals a hemispheric skewness in the distribution of emissions. Aviation activity and emissions are to a large share located in the Northern Hemisphere (92%) and in the areas of the United States, Europe, and Southeast Asia (details in supplementary section C.2).
We calculate emissions for 36.2 million flights in 2019. The global sum of CO2 emissions amounts to 920 Tg, with jet, turboprop, and piston aircraft flights accounting for 99.4%, 0.6%, and less than 0.01%, respectively. With a total distance flown of 
km, we observe average emissions of 82 gCO2 seat-km−1 or 102 gCO2 passenger-km−1, assuming a cabin load factor of 0.8. International flights caused 614 TgCO2 (67%) and domestic flights 306 TgCO2 (33%). The lower domestic share in CO2 emissions than flight numbers (compare supplementary material section C.1) is expected, given the longer average distance of international flights. We provide a short analysis of figures 1(b)–(e) as a case study on Brazilian aviation emissions in the supplementary section C.3.
3.Results
In this section, we first present the resulting national aviation emission inventories. Then, we compare our inventories to currently reported emissions in the UNFCCC database and other national reports. We further benchmark the results at a global scale with other geospatial emissions inventories, showing that results largely align.
3.1.197 national aviation emission inventories
Allocating flights and their emissions to the country of departure allows us to provide aviation emission inventories for all 197 UNFCCC-ratifying countries. Figure 2 shows total (domestic + international) aviation emissions reported under the UNFCCC (a) and calculated by the AviTeam (b). Figure 3 shows corresponding domestic aviation emissions, and supplementary material figure C4 shows international aviation emissions. Whilst under the UNFCCC, a total of 604 TgCO2 are reported for 2019, we calculate a total of 911 TgCO2 emitted by UNFCCC parties in the same year. Another 9 TgCO2 are allocated to non-UNFCCC parties Taiwan, Kosovo, South Sudan, and West Sahara. The difference is caused by a total of 197 countries being quantified in our dataset in contrast to 45 emission reports under the UNFCCC, 43 of them by Annex I countries.
We find emissions of 417 TgCO2 by non-Annex I countries and 494 TgCO2 by Annex I countries. Normalising with population numbers (UN 2023), this gives average emissions of around 364 kgCO2 per capita for Annex I countries and 64 kgCO2 per capita for non-Annex I countries. The complete set of national data, including the year of last reported international aviation emissions under the UNFCCC (UNFCCC 2023a), is presented in the supplementary material tables C4–C7 and for non-CO2 emissions in supplementary material data 1.
Importantly, for many non-Annex I countries, e.g. Bangladesh, Djibouti, etc these may be the first aviation emission inventories beyond CO2 inventories aligned with the UNFCCC 3B methodology. For other countries, e.g. Brazil, these may be the first national inventories of this kind since 2016 (figure 2). In the case of 45 countries that have never reported complete national aviation emissions (domestic + international) under the UNFCCC, we provide estimates of their national aviation emissions in 2019 that sum to 110 TgCO2 (12% of total CO2 for 2019). CO2 emissions from international departures in these 45 countries sum to 93 Tg (84%) and domestic departures to 17 Tg (16%). These shares compare to an international share of 64% and a domestic share of 36% in Annex I countries, confirming our initial assumption that domestic aviation is of subordinate importance in many of these 45 countries.
Ninety-five countries last reported values under the UNFCCC before 2016. Their total emissions in 2019, calculated with AviTeam, sum to 236 TgCO2, and international departures contribute 139 TgCO2. The sum of the last international aviation emissions reported to the UNFCCC in these countries is 73 TgCO2, thus showing growing aviation emissions.
In the 2019 set, the five countries with the highest CO2 emissions from departing flights (global percentage share in total CO2 aviation emissions in parentheses) are the USA (22.2%), China (14.4%), Great Britain (3.8%), Japan (3.5%), and the United Arab Emirates (3.3%) (figure 2). The top five non-Annex I countries (all without 2019 values reported to the UNFCCC) are China (14.4%), the United Arab Emirates (3.3%), India (2.9%), Thailand (1.8%), and Brazil (1.6%). We observe zero aviation emissions for Andorra, Liechtenstein, Monaco, Vatican, and San Marino, as these countries do not possess any commercial airports, given the caveat that helicopter flights are fully and general aviation flights partially excluded from this inventory (figure 2).
The countries with the highest absolute domestic aviation CO2 emissions (global percentage share in total CO2 aviation emissions in parentheses) are the USA (13.4%), China (8.9%), India (1.5%), Russia (1.2%), and Japan (1.1%) (figure 3). Domestic CO2 emissions in kg per capita per year are largest in the USA (366), Australia (252) and Norway (162) (supplementary figure C7).
If international aviation emissions were included in national budgets and NDCs based on the departure country, they would imply limited changes (
) for many of the largest economies (figure 4). However, aviation emissions could substantially change (
) the emission records of other countries, e.g. Ethiopia, Portugal, and Niue. Many of these countries have in common that they are popular destinations for international tourism or host important airports for connecting flights, revealing a limitation to purely departure country-based allocation.
Further, we group emissions by continent (supplementary table C8), and compute emissions that would have been covered by ICAO's CORSIA scheme (international flights that departed and arrived in a country participating in the scheme, status 2023). 8.1 Million flights emitting 336 TgCO2 convert to a CORSIA coverage of 37% of global aviation CO2 emissions based on 2019 values (supplementary table C8).
Aviation emissions per country are compared to other macroeconomic variables (figures and regression analysis in supplementary section C.4). Total aviation emissions show the strongest correlation with the GDP and economy-wide CO2 emissions, domestic aviation emissions also correlate with the territory area. Differences between Annex I and non-Annex I countries' aviation emissions exhibit largely similar behaviours relative to those variables. Linear regression finds (sign of coefficient in parenthesis) GDP (+), GDP per capita (+), Population (+), log(Population) (−), Gini coefficient (+), territory area (+), Asia (+), and railway passengers (−, which can be interpreted as an indicator of the competition of rail and aviation for domestic travel) as good explanatory variables for total aviation emissions.
3.2.Comparison to national and global aviation emission inventories
CO2 emissions from international and domestic departures are compared against UNFCCC inventories (UNFCCC 2023a) and a study by the International council on clean transportation (ICCT) on aviation emissions for 2019 for validation purposes (Graver et al 2020) (figure 5). Our results cover 197 countries, the ICCT reports 106 countries, and the UNFCCC data 45 countries for international and 43 for domestic emissions. The comparison shows a high agreement with UNFCCC values for international aviation (figure 5(a)) with an R2 of 0.99 (ICCT: 0.87), with many of the national inventories showing a difference between our and UNFCCC values of less than 20%. On average, UNFCCC values are 18% larger than emissions quantified with the AviTeam. A likely cause for the deviations is the partial coverage of private and general aviation and the exclusion of helicopter traffic from this work. This becomes relevant for countries with very few or no commercial domestic flights, such as Slovenia (figure 5(b)). The larger share of commercial aviation flights also explains a larger agreement between the AviTeam and the UNFCCC reports in the international segment. Compared to the ICCT, the estimates by the AviTeam are generally larger, reflecting a lower total emissions estimate by the ICCT compared to other studies. For domestic aviation emissions, an R2 of 0.90 (ICCT: 0.99) and a trend of stronger agreement with higher total emissions are observed.
At a global level, values are compared to other global aviation emission inventories and statistics (table 1). A general agreement of this work with other ADS-B-based inventories is evident (Quadros et al 2022, Teoh et al 2023). The number of flights and fuel burn is close to those by Teoh et al (2023), although different scheduled flight information and ADS-B data were used. The larger fuel burn in Quadros et al (2022) is related to considerably larger emissions in the landing-and-take-off phase (supplementary material figure C9). All ADS-B-based inventories also find a similar spatial distribution (supplementary material figures C10 and C11), yet the use of stylised flight trajectories explains larger local concentrations in Quadros et al (2022). This work presents slightly higher emission indices for NOx mass, nvPM mass, and nvPM numbers than other bottom-up inventories. Differences in the emission indices can be caused by different atmospheric humidity, distribution of thrust settings, and in the case of nvPM indices, methods. The aircraft-engine matching may be another source of differences, particularly concerning HC and nvPM emissions, as differences between engines for the same aircraft type reach factors of 100 and 20, respectively (compare Quadros et al 2022, section S8).
Table 1.Comparison of global aviation emission inventories. ∼ shows approximate numbers.
| AviTeam | Quadros et al (2022) | Teoh et al (2023) | Lee et al (2021) | ICCT | ICAO | ||
|---|---|---|---|---|---|---|---|
| Reference | Quadros et al (2022) (FR24) | Teoh et al (2023) | Lee et al (2021) | Graver et al (2020) | ICAO (2023b) | ||
| Year of emissions | 2019 | 2019 | 2019 | 2019 | 2019 | 2019 | |
| Source | ADS-B and schedule | Enriched ADS-B | ADS-B and schedule | Various | Schedule | Reports | |
| Seat-km |  km | 1.07 | ∼1 | 1.09 | |||
| RPK |  km | 0.84 | ∼0.8 | 0.87 | 0.86 | ||
| Flights | Million | 36.2 | 36.5 | 38.8 | 38 | ||
| Fuel | Tg | 291 | 297 | 283 | 327 | ||
| CO2 | Tg | 920 | 937 | 893 | 1033 | 785 | |
| SOx | Gg | 174 | 178 | 170 | |||
| H2O | Tg | 358 | 367 | 348 | |||
| NOx | Tg | 5.51 | 4.62 | 4.49 | |||
| CO | Gg | 604 | 814 | 400 | |||
| HC | Gg | 36.3 | 42.6 | 33.9 | |||
| nvPMm | Gg | 20.8 | 9.7 | 21.4 | |||
| nvPMn |  | 3.76 | 3.47 | 2.83 | |||
| EI NOx | g kg−1 | 18.9 | 15.6 | 15.9 | 15.1 | ||
| EI CO | g kg−1 | 2.08 | 2.74 | 1.42 | |||
| EI HC | g kg−1 | 0.125 | 0.143 | 0.120 | |||
| EI nvPMm | g kg−1 | 0.071 | 0.033 | 0.068 | 0.030 | ||
| EI nvPMn (v) |  kg−1 | 1.29 | 1.17 | 1.00 | 0.20 | ||
a RPK: Revenue passenger kilometres. b in S mass. c in NO2 mass. d Calculated from emission indices in Teoh et al (2023), table 1. e nvPMn is the number of non-volatile particulate matter particles.
4.Discussion
Using the AviTeam model, which leverages ADS-B telemetry data, allows us to compile gridded and aggregated inventories for all 197 UNFCCC-ratifying countries. The compiled set of inventories is aligned with the UNFCCC Tier 3B methodology outline on inventory creation and thus allows comparing newly calculated annual emissions with reported values. Our current inventories focus on the main aviation pollutants. To fully comply with UNFCCC reporting, additional assumptions for CH4 and NO2 emissions and alignment with IPCC quality assurance and verification protocols, as presented in IPCC (2006), are needed. Nonetheless, this set of inventories can be useful in cases where no recent national aviation emission inventory exists and supports previous reports under the UNFCCC.
Taking a share of aviation in total anthropogenic CO2 emissions of 2.4%, our analysis indicates that about 1% of anthropogenic CO2 emissions attributable to non-Annex I countries' aviation were not reported under the UNFCCC for 2019. In light of the expected growth of aviation overall, this supports efforts to extend the emission reporting in non-Annex I countries. A caveat is that while the list of non-Annex I countries includes some large emitters, such as China (values compared against values from Zhang et al (2022) and official records from the CAAC (2020)) and India, it comprises mostly countries where aviation emissions are of smaller absolute values (cf, supplementary material figure C8). The benchmarking further highlights a larger divergence between inventories for smaller countries, potentially motivated by different approaches to handling general aviation and a varying coverage of ADS-B in particular in some developing countries (Teoh et al 2023, SI, p 4 and Quadros et al 2022, S14). Thus, there is value in complementing ADS-B information with country-specific flight records from national agencies that provide even higher completeness than achievable with global commercial flight data (e.g. Klenner et al 2022).
Agreement on international aviation emissions is larger, and these emissions can significantly impact national budgets. For instance, in many island nations or countries of large international airlines aviation emissions, including international aviation, could alter the national total by more than 8%, potentially giving rise to questions regarding the justness of an allocation of emissions to the flights' departure country. Yet, including international aviation emissions into national accounts could reduce the ambiguity in responsibilities (Jaramillo et al 2022).
Our results indicate the largest explanatory value of GDP in predicting the total national aviation emissions but also an explanatory value of other variables, e.g. rail passengers and the Gini coefficient measuring economic inequality. Expecting further GDP growth and continued efficiency increases in the aviation sector, the activity and, hence, emissions from the sector are likely to grow without further mitigation.
The comparison with previous global aviation inventories indicates a general agreement with regard to other emission species and fuel burn, but all are subject to various uncertainties. The largest deviation between the inventories exists in the case of nvPM mass and numbers. This may be caused by the intrinsic difficulty of modelling those emissions in the presence of significant variability of emission indices across engines, engine thrust settings for the same engine, and varying engine data availability (Quadros et al 2022, Teoh et al 2023). Our inventory shows 15% higher NOx emissions, which may be caused by a slightly lower average cruise altitude and more data points close to maximum thrust settings than in other inventories or a different fleet composition. Future inventories may potentially present larger NOx emissions, as Quadros et al (2022),observe a general trend of increasing NOx emissions per kg-fuel over the last decades. Further uncertainty in regard to NOx , HC, CO, and nvPM emission modelling is introduced by relying on emission indices of likely new engines reported to the engine databases, while engine emission indices vary with the engine's operational age (Zaporozhets and Synylo 2017).
Beyond the aggregated data, we generate gridded data consistent with the UNFCCC methodology that may facilitate climate and air pollution simulations that emphasise national or domestic aviation or regional studies. We hope that the inventories may improve the modelling of local aviation effects, strengthen the credibility of existing inventories, emphasise the need for ambitious mitigation targets for domestic aviation, and further the explicit inclusion of aviation into more NDCs. Stronger mitigation efforts in domestic aviation potentially spill over to international aviation, eventually aiding in curbing the climatic effects of the entire aviation sector.
Acknowledgments
We thank AVINOR A S for providing flight registers for Norway, Cirium aviation analytics for scheduled flight data, and ADSBexchange.com, LLC for providing ADS-B data. We further acknowledge the role of Sigma2—the National Infrastructure for High Performance Computing and Data Storage in Norway (Projectnr. NS9576K, nn9576k), and the Industrial Ecology Digital Laboratory, as providers of computing and storage infrastructure for our simulations. We thank Eurocontrol BADA and the anonymous reviewers for valuable feedback.
This document has been created with or contains elements of Base of Aircraft Data (BADA) Family 3 Release 3.15 which has been made available by EUROCONTROL to the Norwegian University of Science and Technology. EUROCONTROL has all relevant rights to BADA. ©2019 The European Organisation for the Safety of Air Navigation (EUROCONTROL). All rights reserved. EUROCONTROL shall not be liable for any direct, indirect, incidental or consequential damages arising out of or in connection with this product or document, including with respect to the use of BADA.
Data availability statement
The data that support the findings of this study are available in the supplementary material and additional data are available upon reasonable request from the authors.
Funding
The authors of this work received funding from the Research Council Norway-funded Project FME NTRANS (Grant No. 296 205).
CRediT role(s)
J K: Conceptualization, Methodology, Software, Formal analysis, Writing—original draft; H M: Conceptualization, Analysis, Resources, Writing—review & editing; A H S: Conceptualization, Supervision, Writing—review & editing.
