Carbon Calculations for An EU-Wide Train Over Plane Zone

With air travel accounting for 2.5% of global CO2 emissions, a number of EU countries are phasing-out certain domestic flight routes in favour of greener rail alternatives.

As pledges of ‘green goodwill’ from COP26 fade into ancient history, the run-up to COP27 raises yet more concerns over the disconnect between intentions and actions.

In light of this, we’ve spent the last few months developing Train Over Plane — an interactive map that models the potential carbon savings associated with global leaders taking a proactive stand against short-haul aviation.

If you haven’t seen it already, hit the button below to start exploring.

Explore Train Over Plane

For those of you who would like to dive deeper into our three-stage model and get to grips with the logic behind our calculations, the following provides a step-by-step walk-through of our methodology.

Disclaimer: Train Over Plane is an ever-evolving tool that aims to keep a live record of how global leaders and private bodies are responding to calls to decarbonise the transport sector. With this in mind, it’s important to recognise the need for more research and more data to improve the accuracy of our calculations and establish a more representative model of how changes in both public policy and passenger behaviour translate into carbon savings.

Methodology for Stage 1

Stage-Wide Rule: Active legislation or ongoing conversations about phasing-out specified flight routes that can be completed by train in less than 2.5 hours.

Country-Specific Rules:

• France: Legislation to replace domestic flight routes that can be completed by a direct high-speed TGV train in under 2.5 hours.

• Austria: Government incentive to replace the Austrian Airlines flight route between Vienna and Salzburg with a train service.

• Spain: Government recommendation to replace all domestic flight routes that can be completed by train in under 2.5 hours.

• The Netherlands: Active partnership between private travel operators to minimise short-haul flight routes to neighbouring countries. There are no domestic flights in the Netherlands and Brussels is one of the closest international airports to Amsterdam.

Stage 1 Calculation

• After identifying France, Austria, Spain and the Netherlands as countries that have either active legislation or ongoing discussions about replacing domestic flight routes with train services, we researched which specified routes fall under our 2.5-hour threshold.

• We then found the aerial distance between each departure/arrival point and used Google Flights to find the number of flights (in both directions) on each day of the week. Note that we used data from the week beginning Monday 25th April 2022 for all Stage 1 calculations, except for routes in France. For Stage 1 routes in France, we used historical flight data sourced prior to the ban that came into effect in April 2022.

• We then found the primary type of aircraft used for each route to estimate the average number of passengers per flight and multiplied this with the estimated number of annual flights. This figure was then multiplied by the ‘proportion of flights terminated’ which we assumed to be 1 (i.e. 100% of the routes listed would be terminated). This figure gave us the annual reduction in air passenger kilometres.

• We’ve also included a metric for ‘Modal Shift’ or the conversion rate of flight-to-train passengers in our calculation. This takes into account passenger attitudes towards train travel and could be adjusted with time as both infrastructural improvements and consumer behaviours adjust to a carbon-conscious future. For the sake of this exercise, we assumed the conversion rate to be 100% but this could be adjusted if we conducted further research on passenger attitudes towards rail across France, Austria, Spain and the Netherlands. Multiplying this figure with the annual reduction in air passenger kilometres provides the rise in annual train passenger kilometres.

• We then multiplied the reduction in air passenger kilometres by the average carbon emissions of a short-haul flight (0.36kg/passenger kilometre) to find the total carbon saving of terminating the specified flight routes.

• Simultaneously, we multiplied the increase in train passenger kilometres by the average carbon emissions of a train (0.0351kg/passenger kilometre) to find the additional carbon emissions caused by replacement train routes.

• Finally, we subtracted the additional train emissions from the reduced air emissions to find the total reduction in carbon emissions.

Methodology for Stage 2

Stage-Wide Rule: Ongoing conversations about replacing specified flight routes that can be completed by train in less than 4 hours.

Country-Specific Rules:

• France: Recommendation from France’s Citizens Convention on Climate to replace domestic flights that can be completed by train in under 4 hours.

• Austria: Extending Austria’s Train Over Plane zone to 4 hours to replace flights between Vienna & Innsbruck, Vienna & Klagenfurt, Linz & Innsbruck, and Salzburg & Graz.

• Spain: Extending Spain’s Train Over Plane zone to include routes between Madrid, Barcelona, Malaga, Seville and Valencia.

• The Netherlands: Extending the Train Over Plane zone to 4 hours by restricting flight routes between Amsterdam and major cities in neighbouring countries.

• Germany: Ongoing conversations about investing in rail infrastructure and replacing flight routes that can be completed by rail in under 4 hours.

Stage 2 Calculation

• After identifying France, Austria, Germany, Spain, and the Netherlands as countries that have either active legislation or ongoing discussions about replacing domestic flight routes with train services, we researched which specified routes fall under our 4-hour threshold.

Note: The reason for including Germany in the Stage 2 calculations and not the Stage 1 calculations is because, although Germany has active/ongoing conversations about phasing-out certain flight routes in favour of rail, the major routes that would be affected fall outside of the 2.5-hour threshold used in Stage 1.

• We then found the aerial distance between each departure/arrival point and used Google Flights to find the number of flights (in both directions) on each day of the week. Note that we used data from the week beginning Monday 25th April 2022 and the week beginning Monday 9th May 2022.

• We then found the primary type of aircraft used for each route to estimate the average number of passengers per flight and multiplied this with the estimated number of annual flights. This figure was then multiplied by the ‘proportion of flights terminated’ which we assumed to be 1 (i.e. 100% of the routes listed would be terminated). This figure gave us the annual reduction in air passenger kilometres.

• We’ve also included a metric for ‘Modal Shift’ or the conversion rate of flight-to-train passengers in our calculation. This takes into account passenger attitudes towards train travel and could be adjusted with time as both infrastructural improvements and consumer behaviours adjust to a carbon-conscious future. For the sake of this exercise, we assumed the conversion rate to be 100% but this could be adjusted if we conducted further research on passenger attitudes towards rail across France, Austria, Germany, Spain and the Netherlands. Multiplying this figure with the annual reduction in air passenger kilometres provides the rise in annual train passenger kilometres.

• We then multiplied the reduction in air passenger kilometres by the average carbon emissions of a short-haul flight (0.36kg/passenger kilometre) to find the total carbon saving of terminating the specified flight routes.

• Simultaneously, we multiplied the increase in train passenger kilometres by the average carbon emissions of a train (0.0351kg/passenger kilometre) to find the additional carbon emissions caused by replacement train routes.

• Finally, we subtracted the additional train emissions from the reduced air emissions to find the total reduction in carbon emissions.

Methodology for Stage 3

Stage-Wide Rule: An EU-wide ban to replace all domestic and intra-EU flight routes that can be completed by train in less than 4 hours.

Stage 3 Calculation

We’ve used data from Eurostat to calculate an estimate of the total number of additional train passengers and the net carbon saving if all 27 member countries of the EU were to replace domestic and intra-EU flights that can be serviced by trains in under 4 hours.

Note that this is an entirely hypothetical exercise at the time of writing.

• 30% of EU Flights are <500km (sourced from this Eurocontrol publication).

• The mean distance of flights in Stage 2 that could be serviced by a train route in <4 hours is 430km.

• The ‘Rail Sophistication Index’ metric refers to an estimated proportion of flight routes <500km that could be serviced by train in <4 hours. We can adjust these figures to account for the varying levels of train infrastructure across different EU countries (i.e. Romania’s rail network differs significantly from the Netherlands’ rail network). While we have used a conservative one-size-fits-all estimate of 75% for the Rail Sophistication Index, a more detailed calculation would use unique figures for each country based on a data-backed assessment of the following three factors: speed, connectivity & frequency. We could also pull data from here to include the lengths of active railway lines in our calculation. As it stands, we don’t have enough visibility to make granular assumptions about rail sophistication across individual EU countries.

• We’ve also included a metric for the conversion rate of flight-to-train passengers in our calculation. This takes into account passenger attitudes towards train travel and could be adjusted with time as both infrastructural improvements and consumer behaviours adjust to a carbon-conscious future. For the sake of this exercise, we assumed the conversion rate to be 100% but this could be adjusted if we conducted further research on passenger attitudes towards rail across the EU.

Scope For Future Research

This initial iteration of Train Over Plane provides a useful tool to consolidate existing conversations and provide ballpark figures to bring attention to the pressing need for a green transport transformation. That said, further research in this space is vital to provide decision-makers with measurable data that can inform calculated social, economic, political and environmental decisions.

While elements of the existing methodology are limited by time and funding restraints, we’ve identified the following areas of interest that future iterations of Train Over Plane could benefit from:

• The datasets for Stages 1 & 2 are taken from flight schedules across March, April and May 2022. Considering the COVID-induced dip in aviation activity, it would be interesting to repeat the data analysis at a time when flight numbers have increased to ‘new normal’ levels.

• Our calculation of a 4.3 million increase in rail passengers for Germany in Stage 2 aligns exactly with the prediction made by the German Aviation Association and Deutsche Bahn. They state that infrastructural rail improvements and the phasing-out of certain flight routes will convert 20% of the country’s total domestic air passengers into rail passengers — totally at 4.3 million. While it’s possible that this alignment is coincidental, understanding the logic behind their 20% calculation would help to validate or nullify our methodology.

• The Stage 3 calculations use very generic data and provide very little granular detail about individual countries. Owing to the size of the EU, the complexity of calculating routes that stretch across borders, and the project’s time constraints, it wasn’t feasible to apply the Stages 1 & 2 methodologies in Stage 3. That said, it would be interesting to apply the country-specific methodology to a handful of countries in Stage 3 to assess how the results compare to our blanket approach and validate the estimations used across all EU countries in Stage 3.

• As previously discussed, Stages 1 & 2 use a ‘Modal Shift’ index to represent the conversion rate of flight passengers into rail passengers as a proportion. For simplicity sake, we assumed the conversion rate to be 100% to model a hypothetical situation where all air passengers who are affected by the Train Over Plane zone choose to use the train as an alternative. While this is an unrealistic assumption in many ways, using a lower figure for Modal Shift would have been equally unrealistic as we need more information about public attitudes towards rail if we are to make any quantitative claims. Specifically, understanding the variation in public attitudes across different EU countries through consumer insights research will be vital to orchestrating a cross-border effort to improve rail uptake and decarbonise the aviation sector.

• As previously discussed, the Stage 3 calculations used a ‘Rail Sophistication Index’. This represents the proportion of rail routes <500km that can be completed by train in <4 hours and can also be serviced by a direct flight. As it stands, we used an EU-wide assumption that 75% of rail routes would fit the above criteria. While we were unable to find any data to support this figure, we felt it would be unrealistic to assume that 100% of rail routes would fit the bill. In reality, we imagine the Rail Sophistication Index would vary significantly across different nations but more research is needed to quantify this assumption. One possibility would be to score each country’s rail infrastructure against certain parameters (e.g. average train speed, route frequency, network connectivity) to build a more representative picture of EU rail.

Go to Train Over Plane

Help Us Decarbonise Travel

If you’d like to take an even deeper dive into our calculations, this spreadsheet shows how we generated estimations for each stage.

Additionally, if you have any specific queries about Train Over Plane or you would like to discuss ideas as to how we could improve our calculations, we’d love to hear from you.

Please contact our Marketing Director, Frederic Kalinke.

e: marketing@silverrailtech.com