Introduction

In recent years, the aviation sector has seen immense growth, with the number of flights and passengers reaching all-time highs before the impact of the global COVID-19 pandemic. Prior to the pandemic, the aviation sector accounted for ~2.4% of annual human-caused CO2 emissions, and it has been responsible for ~4% of the observed human-induced global warming to date1. Airbus forecasts that demand for passenger traffic will grow annually by 3.6% over the next 20 years, with projections indicating it could contribute to a 0.1 °C increase in global temperatures by 20501. The aviation industry’s striking contribution to greenhouse gas (GHG) emissions underscores the urgent need for accurate and reliable methods to quantify the carbon footprint of commercial aviation. Addressing this challenge directly supports progress towards multiple United Nations Sustainable Development Goals (SDGs), including Climate Action (SDG 13), Responsible Consumption and Production (SDG 12), and Industry, Innovation, and Infrastructure (SDG 9)2.

Current carbon calculators of the impact of aviation fall short of what is required on several accounts. Their narrow scope means that they fail to capture the full breadth and depth of emission sources3,4,5. The lack of accuracy in each tool results from shortcomings in capturing user-specific, reputable, up-to-date and consistent data, and the lack of consistency between tools comes from not deploying standardised methodologies4,5,6,7,8,9. These shortcomings result in inconsistent and inaccurate assessments, posing a serious challenge for airlines, policymakers, and passengers, who rely on precise data to make informed decisions aimed at mitigating environmental impacts10,11,12.

One critical shortcoming of many existing carbon calculations is the lack of inclusion of the non-Kyoto impacts, a term we use to refer to substances created by aeroplanes not covered under the Kyoto Protocol, such as nitrogen oxide (NOx), water vapour (H2O), and contrail-induced cloudiness (CiC), that contribute to aviation’s impact on climate13. The non-Kyoto effects of aviation on climate can be substantially greater than the CO2 effects13, and therefore omitting these effects from carbon calculators results in a severe underestimation of the climate impact of a flight14,15. The Intergovernmental Panel for Climate Change suggests using a radiative forcing index (RFI) multiplier, a measure of the importance of aircraft-induced climate change other than that from the release of fossil carbon alone. Effective radiative forcing is a measure of the rapid response of the climate system to external factors16. Contrail cirrus cloud formation produces the largest positive effective radiative forcing, followed by CO2 and NOx, while sulphate aerosols produce cooling13. Lee et al.13 estimated that the Global Warming Potential over a 20-year time frame (GWP20) for contrail cirrus compared to the CO2 impacts (the CO2e/CO2 ratio) to be 2.3 and over 100 years (GWP100) to be 0.63; they estimated the total non-CO2 impacts of aviation compared to the CO2 impacts to be 4.0 (GWP20) and 1.7 (GWP100). Teoh et al.17, however, estimated the GWP20 of contrail cirrus to be 1.1 and the GWP100 to be 0.29. They suggest that the difference between their estimates and those of Lee et al.13 is due to their analysis incorporating more air traffic in the sub-tropics, where persistent contrails are less likely to form. The non-CO2 impacts of aviation on climate are very significant13 and at least twice as large as the CO2 impacts, with large uncertainties. Including contrail-induced emissions in the carbon footprint calculation acknowledges both a challenge and an opportunity. While contrails may not form for every flight, the modularity of the Air Travel Passenger Dynamic Emissions Calculator (ATP-DEC) allows for flexible modelling based on stakeholder choice. ATP-DEC’s dynamic, parameterised model offers a more scientifically robust and comprehensive assessment than simplified methods such as RFI and contrail “bucket” classification18. The bucket approach provides high-level risk categories based solely on contrail patterns. It omits major contributors like NOx and water vapour and lacks sensitivity to flight-specific parameters. By failing to convey actual climate impact, it offers no transparent quantification and risks misleading users into underestimating the true scope of non-Kyoto aviation effects.

Similarly, very few calculators adopt a full life cycle assessment (LCA) perspective. A cradle-to-grave LCA approach must include the production and transportation of fuel (Well-To-Tank (WTT)), fuel burning during flying, take-off and landing (Tank-To-Wake (TTW)), in-flight services, and airport and aircraft life cycles19,20. Following an evaluation of existing carbon calculator tools, this paper outlines a comprehensive LCA-based methodology for the ATP-DEC, a novel calculator that improves the accuracy of the variables commonly used and the scope of those variables either ignored or poorly estimated. Compared to current tools in the market, ATP-DEC i) better raises awareness of the impact of flying and attributes such impact to different sources, ii) more accurately tracks and reports impacts for corporate reporting, iii) feeds more accurate data to carbon offsetting mechanisms21, iv) identifies specific aspects of CO2e that change over time, v) better informs policies such as aviation taxation. By addressing the limitations of existing methods, this research aims to provide a robust framework that can be easily adopted by stakeholders within the aviation industry. The methodology is analysed over 30,000 flights, and all techniques and variables are explained and justified. Finally, the limitations of the model are stated for complete transparency21.

Results

Comparison of ATP-DEC with existing tools

To benchmark the proposed methodology, carbon footprint estimates were compared against four existing calculators: Google Travel Impact Model (TIM), IATA CO2 Connect, ICAO Carbon Emissions Calculator, and MyClimate. The selection of comparison tools was informed by their prevalence in industry practice and public use, reflecting their relative popularity and accessibility. These tools represent a diverse set of methodologies and serve as benchmarks for assessing the methodological advancements offered by ATP-DEC. The estimations for each of the tools are taken from their online calculator websites, with variables kept the same for each calculator where possible. Results are presented in Fig. 1, which visualises the carbon footprint of a flight between Singapore Changi (SIN) and Zurich (ZRH) across four passenger classes—First, Business, Premium Economy, and Economy—using stacked bars for ATP-DEC and single bars for the existing tools.

**Fig. 1: Comparison of air passenger emission footprint calculator outputs by travel class for a flight from SIN to ZRH[30](https://www.nature.com/articles/s43247-025-02847-4#ref-CR30 “McFall, F. Air travel passenger dynamic emissions calculator article figure dataset. figshare. https://doi.org/10.6084/m9.figshare.29852306.v2

(2025).“).**

Bar chart showing the emissions footprint (in kg CO2e for ATP-DEC, TIM, and MyClimate and kg CO2 for IATA and ICAO) for a flight from Singapore (SIN) to Zurich (ZRH) across four passenger classes: first, business, premium economy, and economy. Four blue shades represent existing calculators, from dark to light: TIM, IATA, ICAO, and MyClimate. Stacked coloured bars represent the ATP-DEC results, with individual emissions components: Well-To-Tank (WTT, dark blue), Tank-To-Wake (TTW, green), carry-on luggage (purple), checked luggage (yellow), in-flight services (light blue), airport (teal), aircraft (red), nitrogen oxides (NOx, orange), water vapour (H2O, light brown), and contrail-induced cloudiness (CiC, dark brown). The inset magnifies the First Class ATP-DEC for improved visibility of small components.

Figure 1 highlights a key distinction of the ATP-DEC: the ability to break down emissions into multiple sources for each passenger class: TTW, WTT, carry on and checked luggage, in-flight services, airport and aircraft life cycle emissions, non-Kyoto gases: NOx, H2O and non-Kyoto impact: CiC. In contrast, the results from TIM, IATA, ICAO, and MyClimate are presented as aggregated emissions, limiting their ability to capture detailed contributors to the carbon footprint. A full breakdown of emissions not only allows stakeholders to decide which emissions matter for their requirements but also facilitates targeted environmental action by highlighting key areas for intervention.

ATP-DEC consistently produces higher overall emission estimates compared to existing tools, particularly for premium passenger classes (First and Business). This discrepancy is primarily due to the inclusion of non-Kyoto impacts (NOx, CiC, H2O), which are excluded by TIM, IATA, and ICAO. While MyClimate includes non-CO2 effects using a simplified RFI, ATP-DEC’s dynamic modelling of NOx, H2O, and CiC more accurately reflects their contributions22 (Equation S1–S35). TIM 3.0 introduces contrail warming “buckets” with the emissions footprint, whereby contrail warming impact is categorised into relative impact levels compared to the fuel burn emissions18. A quantitative estimate of the actual climate impact is not provided, making it impossible to compare or aggregate non-CO2 effects in a transparent way.

Additionally, the Historical Adjustment Factors (HAFs) derived from historical flight data further enhance precision by accounting for actual flight deviations and rerouting (Equation S19), which are not reflected in the static assumptions of the other tools. In this case, this real-world adaptability has captured the increased distance and change in mean latitude, increasing the fuel burn and non-Kyoto impacts.

When non-Kyoto impacts are excluded, the ATP-DEC results align more closely with TIM and IATA. However, ATP-DEC still demonstrates superior accuracy due to improved seat class weightings (Equation S10). By incorporating airline and aircraft-specific seating data (acquired from SeatGuru[23](https://www.nature.com/articles/s43247-025-02847-4#ref-CR23 “SeatGuru. Airline seat maps, flights shopping and flight information, https://www.seatguru.com/

(2025).“)), ATP-DEC produces a more realistic split of emissions between passenger classes. This contrasts with the generalised weightings used by other tools, which fail to account for variations in aircraft configurations and premium seating occupancy.

HAF impact comparison

To evaluate the impact of HAFs on carbon footprint estimates, ATP-DEC was compared against post-flight emissions estimates derived from exact flight tracking data. A TIM-based method is implemented in two scenarios for comparison. This analysis was performed for flights from London Heathrow (LHR) and Incheon (ICN) over the course of 2023, as shown in. Again, input parameters have been kept constant apart from the flight route.

The analysis compares four scenarios:

ATP-DEC using HAF (30-day rolling window) (green solid line): adjusts emissions dynamically using historical data trends from the preceding 30 days, accounting for operational deviations. This is a realistic scenario showing how the ATP-DEC operates.

Theoretical ATP-DEC using actual post-flight distance (blue dashed line): estimates emissions using exact post-flight positional flight data. This is unrealistic in practice, but useful for the validation of HAF.

TIM-based method scenario (red solid line): simulates a normal TIM-based method scenario (with constant distance). This scenario shows how the TIM-based method operates in reality.

Theoretical TIM-based method using actual post-flight distance (orange dashed line): simulates a theoretical TIM-based method scenario by incorporating the actual flight distance to compare with the normal TIM-based method. This is an unrealistic scenario, but useful for validation.

Figure 2 demonstrates that the HAFs allow the output to closely match the emissions trend derived from post-flight data throughout the year. This result highlights the effectiveness of the rolling HAF in capturing real-world variations, such as route deviations, air traffic inefficiencies, and operational changes, without relying on exact post-flight inputs. In this scenario, the flight path is altered due to the closure of Russian airspace, causing an increase in distance, fuel consumption, and emissions. ATP-DEC (green line) captures this trend, but the normal TIM-based method (red line) fails to adapt, consistently underestimating emissions. The dotted yellow line shows the theoretically TIM-based method scenario, where it incorporates the actual post-flight distance. Even in this theoretical scenario, ATP-DEC has a higher estimation. This is due to the extensive scope of ATP-DEC, incorporating factors such as in-flight services, aircraft and airport factors. The disparity between the two TIM-based scenarios is evident, with the method lacking actual distance data underreporting per passenger emissions by 23150 tonnes CO2e across the 473 analysed flights in 2023. In contrast, ATP-DEC, which incorporates a 30-day rolling average HAF, exhibits a slight overestimation by just 51 tonnes CO2e.

**Fig. 2: Emission footprint method comparison for LHR-ICN flights in 2023 without non-Kyoto impacts[30](https://www.nature.com/articles/s43247-025-02847-4#ref-CR30 “McFall, F. Air travel passenger dynamic emissions calculator article figure dataset. figshare. https://doi.org/10.6084/m9.figshare.29852306.v2

(2025).“).**

Line plot showing total flight emissions footprint (kg CO2e) for four methodological scenarios: ATP-DEC using historical adjustment factors (HAF, 30-day rolling window; green solid line), ATP-DEC using actual post-flight distance (blue dashed line), TIM-based method using constant distance (red solid line), and TIM-based method using actual post-flight distance (orange dashed line). The ATP-DEC HAF approach closely matches the emissions trend from post-flight data, capturing operational variations such as rerouting due to Russian airspace closure. The TIM-based constant-distance method consistently underestimates emissions, with a total underreporting of 23150 tonnes CO2e across 473 flights. In contrast, ATP-DEC with HAF slightly overestimates by just 51 tonnes CO2e.

Accounting for changes in-flight route using HAF

To demonstrate the impact of HAFs on capturing real-world flight variations, a case study was conducted on British Airways (BA) flights between London Heathrow (LHR) and Shanghai Pudong (PVG) for the years 2019 and 2023. These flights provide a clear example of how external events, such as the closure of Russian airspace, influence flight routes, distances, and carbon emissions.

Case Study Setup:

2019 Flights: Represent pre-pandemic operations without major disruptions. However, even in this period, flights show minor variations in flight distances and mean latitudes due to operational constraints and minor route inefficiencies.

2023 Flights: BA resumed LHR-PVG flights in April 2023 after a pandemic-related suspension. Upon resumption, flights were rerouted to avoid Russian airspace due to the ongoing conflict, resulting in considerably longer flight paths and altered mean latitudes.

The analysis compares carbon footprint estimates using ATP-DEC with and without HAF, alongside post-flight records. Figures 3 and 4 illustrate the changes in-flight distance and mean latitude, while Fig. 5 demonstrates the resulting impact on carbon footprints for economy-class passengers. Figure 3 demonstrates that actual flight distances in 2023 were substantially longer compared to 2019 due to the rerouting required to avoid Russian airspace. Despite the absence of such disruptions in 2019, those flights also exhibit higher distances than the great circle distance (shown as the red line) (Equation S17). Static methods that rely solely on great circle distances fail to capture these variations. By incorporating historical flight path data, the HAF successfully adjusts for these factors, enabling the ATP-DEC to produce more realistic and precise distance estimates (Equation S8).

**Fig. 3: Actual and great circle distances for British Airways flights between LHR and PVG in 2019 and 2023[30](https://www.nature.com/articles/s43247-025-02847-4#ref-CR30 “McFall, F. Air travel passenger dynamic emissions calculator article figure dataset. figshare. https://doi.org/10.6084/m9.figshare.29852306.v2

(2025).“).**

Monthly mean flight distances (km) for 2019 (orange circles and line with shaded range) and 2023 (blue circles and line with shaded range) compared with the great circle distance (red solid line). Shaded areas represent the monthly range of distances. The 2023 flights, resumed in April after the pandemic, were rerouted to avoid Russian airspace, resulting in longer distances than in 2019. Even in 2019, actual distances exceeded the great circle distance due to operational constraints.

**Fig. 4: Mean latitude for British Airways flights between LHR and PVG in 2019 and 2023[30](https://www.nature.com/articles/s43247-025-02847-4#ref-CR30 “McFall, F. Air travel passenger dynamic emissions calculator article figure dataset. figshare. https://doi.org/10.6084/m9.figshare.29852306.v2

(2025).“).**

Monthly mean flight latitudes (°) for 2019 (orange circles and line with shaded range) and 2023 (blue circles and line with shaded range) compared with the mean latitude of departure and arrival airports (red solid line). Shaded areas represent the monthly range of latitudes. In 2023, rerouting to avoid Russian airspace resulted in a lower mean latitude compared to 2019, reflecting a southward deviation in flight paths. Even in 2019, actual latitudes varied month to month due to operational adjustments.

**Fig. 5: Emission footprint estimates for economy-class passengers on British Airways flights between LHR and PVG in 2019 and 2023[30](https://www.nature.com/articles/s43247-025-02847-4#ref-CR30 “McFall, F. Air travel passenger dynamic emissions calculator article figure dataset. figshare. https://doi.org/10.6084/m9.figshare.29852306.v2

(2025).“).**

Monthly emissions footprint (kg CO2e) estimated by ATP-DEC using historical adjustment factors (HAF; 2019: orange circles and line, 2023: green circles and line), ATP-DEC based on post-flight records (2019: purple circles and line, 2023: blue circles and line), and ATP-DEC without HAF (red solid line). Shaded areas represent the monthly range of values. In 2023, rerouting to avoid Russian airspace led to higher emissions compared to 2019. The HAF approach closely matched post-flight estimates in both years, whereas the static-distance method without HAF consistently underestimated emissions.

Changes in mean latitude, shown in Fig. 4, further highlight the impact of rerouted flights. The 2023 flights, which required a southward deviation to bypass Russian airspace, exhibited a reduced mean latitude compared to 2019. These shifts are particularly relevant for calculating non-Kyoto impacts, as atmospheric conditions vary by latitude, influencing the formation of NOx, contrails, and water vapour (Equation S1–S3). Even for 2019 flights, where rerouting was not a factor, variations in mean latitude were observed due to operational adjustments. The HAF dynamically captures these changes, ensuring that latitude-dependent emissions estimates reflect real-world conditions rather than static assumptions, such as using the mean latitude of airports (red line).

The results clearly show the ability of HAF to account for changes in-flight behaviour and improve the accuracy of emissions estimates. The impact of these changes on emissions is evident in Fig. 5. The longer distances and altered latitudes in 2023 resulted in a higher carbon footprint for economy-class passengers, as seen in the post-flight estimates (blue line). Using HAF (green line) aligns the estimations with these post-flight values, demonstrating its ability to account for both large-scale disruptions and routine operational variations. In contrast, ATP-DEC without HAF (red line) consistently underestimates emissions by assuming static flight distances and latitudes. Even for 2019, the HAF enhances the accuracy of the ATP-DEC by capturing subtle variations in route efficiency, which static models overlook.

Analysis of HAF rolling window length

The impact of the HAF rolling window length was assessed using 7-day and 30-day windows. Table 1 summarises the mean squared percentage error (MSPE) for each flight route across the dataset, comparing pre-adjustment results with post-adjustment performance.

The pre-adjustment method, which does not incorporate any historical data, consistently produces the highest MSPE values. This highlights the substantial estimation errors introduced when historical trends in route variability, delays, and operational adjustments are ignored.

The results indicate that the 7-day rolling window generally achieves the lowest MSPE values, providing superior short-term precision. This is particularly evident for flight routes with high variability, where a shorter window more effectively captures recent operational trends, such as sudden changes in delays or rerouting. For these routes, the 7-day HAF quickly adapts to fluctuations, reducing estimation errors.

By contrast, the 30-day rolling window shows slightly higher MSPE values for short-term variability but demonstrates competitive performance, especially for routes exhibiting more stable trends or pronounced seasonality. For the cleaner and less noisy 2023 flight data, the 30-day window even marginally outperformed the 7-day window over the full year. This suggests that, for long-term trends and less-volatile routes, a longer window can smooth transient anomalies and offer more precise results.

Discussion

Evaluation of ATP-DEC against existing tools

Table 2 presents a condensed comparative analysis of the ATP-DEC against TIM, IATA, ICAO, and MyClimate. The comparison is structured around the four core evaluative criteria: scope, consistency and accuracy, transparency, and effectiveness of communication (Table 2).

Strengths and innovation of ATP-DEC

Granularity of emissions breakdown

One of the defining strengths of ATP-DEC is its ability to decompose emissions into distinct sources, such as TTW, WTT, carry-on and checked luggage, non-Kyoto impacts (NOx, H2O, CiC), and life cycle emissions from in-flight services, airports, and aircraft. This allows users to pinpoint emission sources and prioritise effective interventions. By offering a comprehensive view of aviation emissions, ATP-DEC supports both informed decision-making and targeted environmental action. ATP-DEC’s modular framework allows users to include or exclude specific variables; the tool is adaptable to a variety of contexts.

Furthermore, ATP-DEC is modular, so it’s flexible to evolving science and data. Each individual constituent of ATP-DEC can be reworked without affecting the others. This also encourages advancement. For example, the default value for fuel emission factors is Jet A1 fuel. But this can be easily altered when SAF or electric aircraft are used in the future.

Inclusion of non-Kyoto impacts

By incorporating non-Kyoto effects, the ATP-DEC provides a more comprehensive assessment of aviation’s climate impact. This includes dynamic modelling of NOx, H2O, and CiC, which are often overlooked or oversimplified in existing calculators. The ability to accurately estimate these emissions aligns the ATP-DEC with evolving scientific understanding, filling a critical gap in current methodologies15.

Real-world adaptability

Through the integration of HAFs, ATP-DEC captures real-world variations in-flight operations, such as rerouting due to airspace closures or delays. Static models lack this adaptability. The inclusion of rolling windows for HAF further enhances its ability to track both short-term fluctuations and long-term trends.

Validation of this approach was carried out using a dataset of operational flight records that included the actual flown distance and the actual mean latitude for each flight. For each day in the dataset, we compared the estimated GCD and estimated mean latitude (calculated purely from origin-destination coordinates) with the corresponding observed values. From these comparisons, an HAF was computed for both distance and mean latitude by taking the ratio of actual to estimated values of each flight. For example, a distance HAF of 1.05 indicates that actual flight paths are ~5% longer than the direct geographic estimate due to routing constraints, weather avoidance, or air traffic control. The adjustment factors were then applied to the estimates of subsequent flights to better approximate real operational conditions. To stabilise the correction and reduce the impact of outlier days, the HAFs were computed using rolling windows of either 7 or 30 days, ensuring they remained responsive to recent trends while retaining generalisability. The results indicated that the 7-day window was more accurate and will be adopted as the default for the model going forward.

Improved passenger class weighting

The ATP-DEC’s approach to seat class weighting represents another improvement over existing tools. By utilising airline and aircraft-specific seating data, the ATP-DEC ensures fair and realistic attribution of emissions.

Transparency and actionability

The transparency of the ATP-DEC enhances its utility for stakeholders. By providing detailed outputs that clearly attribute emissions to specific sources, the tool facilitates accountability and supports actionable insights. The publication of the methodology allows stakeholders to fully understand their footprint.

Technological integration and scalability

The ATP-DEC methodology is designed for seamless technological integration and scalability, ensuring its applicability across diverse operational contexts. The required input data is already widely available through existing industry sources. Furthermore, its modular design allows adaptation when certain data variables may be missing or incomplete. This adaptability ensures that the tool remains functional, making it particularly useful for regions or sectors with variable data availability.

The deployment of the ATP-DEC is straightforward, with its architecture allowing easy integration into existing software systems, such as airline booking platforms or carbon offsetting programs. Its design supports scalability, enabling implementation across multiple flights or operational networks without much computational or logistical overhead.

A key feature of the ATP-DEC is its symbiotic relationship with a blockchain-based carbon offsetting mechanism. This integration facilitates transparent and verifiable environmental action, allowing stakeholders to directly link emissions calculations with projects with a documented positive environmental impact. By leveraging blockchain technology, the methodology ensures traceability and accountability, enhancing trust among stakeholders. The combination of accurate emissions assessments and seamless offsetting ecosystem positions the ATP-DEC as an essential tool for aligning aviation practices with global climate goals.

Model limitations and challenges

HAF forecast

The HAF provides dynamic emissions adjustments based on historical trends, making it uniquely suited for real-time applications. However, for flights booked in advance, the HAF is calculated on the day of booking using the most recent data available. This ensures that the emissions estimate reflects the latest operational conditions, such as route deviations or airspace restrictions at the time of booking. The HAF is not yet suitable for forecasting future flight conditions.

Engine efficiency

The methodology currently considers aircraft type and age, but does not account for the specific engines installed on each aircraft. Since engines are often replaced or upgraded independently of the airframe, this omission may affect the accuracy of fuel burn and emissions estimates, particularly for older aircraft retrofitted with more efficient engines.

Exclusion of soot and sulphate aerosols

While the ATP-DEC’s non-Kyoto impacts modelling incorporates contrail-induced cloudiness (CiC), it does not explicitly account for the impacts of soot and sulphate aerosols. Although sulphate aerosols are thought to exert a cooling effect, their exclusion, along with soot’s warming contributions, reflects a limitation in the Dahlmann et al. method used in the ATP-DEC22. While soot is a contributor to global warming, sulphate aerosols have a cooling effect. With comparable magnitude, sulphate aerosols help to offset soot’s warming effect. Their modelling uncertainty has been recognised in the literature24.

Non-Kyoto impacts uncertainty

The scientific understanding and modelling of non-Kyoto impacts is still rapidly evolving. While the Dahlmann et al. method was developed based on the emission profile of A330-200, the most commonly sold medium and long-range aircraft22, other aircraft could produce different results. Dahlmann et al.20 calculate that their method produces a mean squared error of 0.19 compared to advanced contrail modelling of specific flights using detailed flight data, whereas using a constant multiplier to assess non-Kyoto emissions produces a mean squared error of 1.18 for these flights. Dahlmann et al.’s22 method remains the most credible for estimating non-Kyoto impacts in terms of CO2e but is constrained by the available scientific data.

Scalability of contrail modelling

Advanced contrail modelling solutions exist that incorporate vast amounts of real-time atmospheric data, offering higher precision. However, these methods are computationally intensive and unfeasible for flight-by-flight calculations at scale. As a result, the ATP-DEC adopts a practical approach that balances accuracy with computational efficiency, recognising that more granular contrail modelling would require considerable technological advancements.

Sensitivity of variables

A sensitivity analysis was outside the scope of work for this study, but it’s acknowledged as an important aspect of future work. Evaluating how variations in key input parameters affect model output can help identify dominant variables and better characterise model uncertainty. Various methods can be used, including both deterministic and probabilistic single or multi-variable approaches. Data uncertainty can be managed by probabilistic or stochastic approaches, such as Monte Carlo simulation and some deterministic ways, such as scenario analysis and sensitivity analysis25.

Conclusions

The ATP-DEC methodology represents a substantial advancement in aviation carbon footprint calculation, offering superior accuracy, scope, and real-world applicability. It sets a new benchmark for addressing the complex challenges of emissions accountability while fostering transparency and actionable insights.

Superior methodology

By incorporating advanced modelling techniques such as HAFs, which utilise historical flight data to achieve exceptional accuracy. Unlike existing tools, it includes an expansive scope that covers the full life cycle of emissions. This footprint breakdown allows stakeholders to understand their environmental impact in a more meaningful way, promoting effective communication and targeted interventions. Furthermore, the transparency of the methodology, demonstrated through its detailed publication, ensures trust and credibility.

Benchmark for future standards

Current standards, such as ISO 14083, remain constrained by limited scope and accuracy. However, the data required to dramatically improve these estimations already exists, as demonstrated by ATP-DEC’s ability to match real-world trends. This methodology not only highlights the inadequacies of existing frameworks but also sets a precedent for the adoption of more comprehensive and scientifically informed standards.

Nudge to environmental action

ATP-DEC is designed to drive environmental action. Its ease of deployment ensures that it can be readily integrated into existing operational systems, leveraging widely available data. This accessibility makes it a valuable component of broader sustainability strategies. While the ATP-DEC itself is not the sole solution to aviation emissions, it functions as a critical part of the ecosystem, working symbiotically with initiatives such as blockchain-based carbon offsetting mechanisms21. This integration facilitates traceable and verifiable environmental actions, empowering stakeholders to engage in meaningful climate mitigation efforts.

In conclusion, the ATP-DEC methodology represents a transformative approach to aviation emissions accounting. By combining superior accuracy, extensive scope, and actionable insights, it surpasses existing tools while setting a foundation for future standards. Its potential to inspire confidence, inform policy, and drive sustainable practices underscores its critical role in addressing the aviation industry’s contribution to global climate goals.

Methods

Approach overview

ATP-DEC follows an LCA approach in line with ISO 14083, quantifying GHG emissions per passenger for a given flight. The cradle-to-grave system scope in the aviation sector’s context includes the production and transportation of fuel WTT, fuel burning during flying, take-off and landing TTW, in-flight services, and airport and aircraft life cycles (Table S1). In addition, ATP-DEC incorporates non-Kyoto impacts specific to aviation. Emission allocation incorporates luggage emissions, aircraft-specific seating configurations, passenger load factors, and cargo load factors. The goal of the tool is to ensure no source of climate impact is overlooked while dynamic models, specific inputs, and life cycle data are combined to deliver high accuracy. HAFs refine pre-flight estimates by integrating real-world variations.

Interpretation: ATP-DEC’s step-by-step calculations

This section summarises the step-by-step calculations to interpret the results, specifically the hotspot analysis of a flight from Singapore to Zurich (Fig. 1). Equation 1 summarises the methodology, with variables described in Table S5. Further justification of each individual variable and how it’s calculated can be found in Section 1.2 of the Supplementary Information (SI).

$${{{{\rm{CO}}}}}_{2}{{{\rm{e}}}} , {{{\rm{per}}}} ; {{{\rm{pax}}}} = {{{\mathcal{G}}}}\left(\frac{{{f}}({{{\rm{D}}}})\times ({{{\rm{WTT}}}}+{{{\rm{TTW}}}})\times {{{\rm{DT}}}}\times {{{\rm{CF}}}}\times {{{\rm{LF}}}}\times {{{\rm{CW}}}}}{{{{\rm{PLF}}}}}\right) \ + L+\left(V\times {{{\rm{IFS}}}}\right)+{{{\rm{APF}}}}+{{{\rm{AF}}}}+\left({{{{\mathcal{Z}}}}}_{1}\times {{{\rm{N}}}}({{{{\mathcal{Z}}}}}_{2})\right)$$

(1)

For steps 1–6 the scope aligns with ISO 14083 and provides a robust, standardised framework for emission calculations. The extended methodology, 7–9, supplements the method, elevating comprehensiveness. The additional and enhanced factors within the extended methodology go beyond the scope of ISO 14083 and are optional. The ATP-DEC methodology outlines a full, holistic, modular formula that acknowledges the importance of a thorough environmental assessment while allowing flexibility for the stakeholder to encourage environmental action.

Calculate distance

Calculate the initial distance using the ({\mbox{GCD}}) and ({\mbox{HAF}}) (Equation S8).

For Fig. 1 (Singapore to Zurich), the ({\mbox{GCD}}) is 10,309 km, calculated using the geographical coordinates of the airports. An HAF can be used to adjust this to a more precise flight distance (Equation S13). For Fig. 1, historical flight data from the arbitrarily chosen final week in December 2019 were used to calculate a distance HAF of 1.0315 (Equation S12 and S19). This means that the average distance of a flight in this month was 1.0315 times higher than the ({\mbox{GCD}}), and we use this to predict the distance of the flight example in Fig. 1 as 10,634 km. The HAF can be updated with a rolling window of 7 or 30 days. For Fig. 1 a rolling window of 7 is used because it proved to be the most accurate in capturing trends (Table 1).

Extract fuel burn

Use a polynomial regression model alongside the fuel burn data to equate the distance flown with the aircraft-specific fuel consumption. Include both the landing, taxi and take-off (LTO) and cruise, climb, descent (CCD) fuel consumption. Multiply by age deterioration factor to account for aircraft efficiency[26](https://www.nature.com/articles/s43247-025-02847-4#ref-CR26 “Cirium. Cirium Emissions Methodology: powered by EmeraldSky (Version 1.1), https://assets.fta.cirium.com/wp-content/uploads/2024/02/09213359/Cirium-Emissions-Methodology-V1dot1-2022-EmeraldSKy.pdf

(2024).“). Where exact fuel burn is known, this value should be used instead.

For the results in this study, Eurocontrol fuel data is used27. Fuel burn for 10,634 km for a Boeing 777–300 (B773) is 89,540 kg (Fig. S1). For a wide-bodied B773 with an age of 1 year, the deterioration factor is 1.01[26](https://www.nature.com/articles/s43247-025-02847-4#ref-CR26 “Cirium. Cirium Emissions Methodology: powered by EmeraldSky (Version 1.1), https://assets.fta.cirium.com/wp-content/uploads/2024/02/09213359/Cirium-Emissions-Methodology-V1dot1-2022-EmeraldSKy.pdf

(2024).“), so the fuel consumption is given as 90,435 kg (Fig. 1).

Compute total aircraft emissions

Multiply the fuel consumption by the WTT and TTW emissions factors and sum them together to calculate the total fuel emissions for the flight. This includes all passengers, passenger luggage, and cargo emissions.

The WTT and TTW emissions for 90,435 kg of Jet A1 fuel are 43,409 kg CO2e and 28,5776 kg CO2e, respectively. This is calculated using emission factors from ISO 1408328 (Equation S20).

Adju