9 minute read.A second look at Crossref’s carbon footprint - the 2024 report
In 2022, we wrote a blog post “Rethinking staff travel, meetings, and events” outlining our new approach to staff travel, meetings, and events with the goal of not going back to ‘normal’ after the pandemic and said that in the future we would report on our efforts to balance online and virtual events, work life balance for staff, and track our carbon emissions. In December 2024, we wrote a blog post, “Summary of the environmental impact of Crossref,” that gave an overview of 2023 and provided the first report on our carbon emissions. Our report on 2023 only just made it into 2024, so we are happy to report on 2024 a little sooner in the year.
On the positive side, there are a few things:
- Our spending on travel and meetings (a proxy for emissions) in 2024 was 56% of what it was in 2019, keeping below the target of not more than 60% of our 2019 spend
- We were better at tracking hotel nights in 2024 compared to 2023
- We managed to balance in-person, regional, and online meetings to engage with our global community while still not having returned to the pre-pandemic “normal”
In practice, our approach means thinking carefully about how to make the most of each trip. For example, when organising our Crossref Jakarta event, we travelled via Singapore and used the opportunity to meet with members there. Once in Jakarta, we combined our two-day event with an OJS workshop with colleagues from PKP, and another event with Universitas Indonesia. Similarly, when our colleague travelled by train to a conference in Amsterdam, they combined it with a day of visits to members in the area. These kinds of combinations reduce the need for separate trips and maximise the value of in-person travel.
Some of the less positive things were:
- As our membership continues to grow globally and we expand our staff, (which are both great things in themselves), our emissions have also increased. Not only do we have more staff, but some staff travelled more in 2024 than in 2023. We’ll keep a close eye on this to avoid ever-increasing travel.
- Taking a train instead of flying can take longer, and clashes with our desire for staff needing to be away from home as little time as possible.
- It is difficult to find reliable data for some calculations - for example, we have decided not to try to calculate the impact of our Zoom use because there is no reliable way to do this.
- We don’t have good options for offsetting our emissions, and it’s unclear whether we would want to do this even if they were available.
There is also the issue of whether it is worth it, or possible, to collect certain data, or whether it would change what we do. An example is Zoom. The estimate for the emissions from Zoom meetings in 2024 was 100 kg (that’s kilograms, not tonnes), but the calculations were made using a tool from 2020 that made many assumptions and estimates. We have no way of verifying whether the tool we used is accurate, so we decided not to update our previous calculation. In any case, we aren’t going to ration or reduce our teleconferencing, since it’s an essential tool, and especially if we want to fly less, have fewer in-person meetings, and operate effectively as a distributed organisation in multiple countries with no offices.
In summary, our total reported carbon emissions increased 40% from 105 tCO2e in 2023 to 147 tCO2e in 2024 (see below for the details). The positive aspect of this is that the increase is partly due to our improved ability to track our travel and hotel stays. The more concerning side of this is that we are travelling more. This enables us to engage with our growing community. We are still thinking strategically about our travel and meetings, following the approach outlined in our 2022 blog post. However, we need to carefully consider air travel in 2026, as it is our largest source of emissions (93%).
Total travel and carbon spending
| Year | Amount | Percentage of 2019 | Total carbon spent | Total hotel nights covered |
|---|
| 2019 actuals | $585,482 | 100% | did not record | did not record |
| 2020 actuals | $91,700 | 16% | did not record | did not record |
| 2021 actuals | $19,066 | 3% | did not record | did not record |
| 2022 actuals | $74,416 | 13% | did not record | did not record |
| 2023 actuals | $305,737 | 52% | 105 tCO2e | did not record |
| 2024 actuals | $327,939 | 56% | 147 tCO2e | 415 |
| 2025 budget | $417,767 (reforecast) | 71% | 68 tCO2e (YTD) | 256 (YTD) |
| 2026 budget | $439,817 | 75% | TBD | TBD |
In 2024, we met the target of keeping our travel expenses below 60% of our 2019 level. In 2025, we will exceed this. There are a number of reasons for this. We have more staff, more members, inflation has been high, and we are subsidising a lot more travel for others, such as our ambassadors, speakers, and collaborators at local events, and some board members (since 2019, we reduced from three to one in-person board meeting per year). This aligns with our goals of inclusivity for Crossref meetings, but we have to recognise there is a trade-off. The cost of travel, particularly airfare, has increased since 2019. Using US Bureau of Labor Statistics data from 2019 to 2025 the inflation multiplier for a dollar is 1.26 so adjusted for inflation the comparison figure for 2025 spending is $737,000 and forecasted 2025 spending is 60% of this. While we use cost as a proxy for travel volume, now that we’re better at tracking actual carbon emissions, we can try to set targets of keeping under a certain carbon tonne equivalent total instead of (only) a financial target.
Total Carbon Emissions for 2024
Our total reported carbon emissions increased 40% from 105 tCO2e in 2023 to 147 tCO2e in 2024. In 2023, we didn’t report on the estimated emissions from hotel stays, but for 2024, we have. We recorded 415 hotel nights in 2024 for 4 tCO2e using an average of Europe/US hotel per night emissions estimates (Circular Ecology). The most carbon-intensive activity was flying. There were about 215 flights in 2024, accounting for emissions of 138 tCO2e - 93% of our total. Crossref staff and community members we covered took 88 train journeys with carbon emissions of .47 tCO2e - so the more travel by train, the better, but this isn’t always possible or feasible. We haven’t included estimates of the impact of home working (Crossref is fully distributed), but we have an initial estimate below and will look to improve this analysis for the 2025 analysis and going forward.
Crossref is fully distributed with staff in 11 countries. We used Claude from Anthropic to calculate the emissions from home working for our staff in 2024 and asked for sources to be cited. It provided some approaches for how to go about the calculations but the results were not reliable - for our 46 staff in 10 countries (this is for 2024 - we now have 49 staff in 11 countries) estimates ranged from 5 tCO2e to 28 tCO2e depending on various assumptions such as whether to account for the grid intensity of the countries where staff are based (Our World in Data has grid intensity figures) and what estimate is used for the amount of energy an employee working from home uses each day. Circular Ecology uses UK DEFRA figures to come up with 2.67 kgCO2e/day for home working. So a simple calculation of 46 staff working 230 days per year arrives at the 28 tCO2e amount. This is much less than the equivalent figure for office-based work, which is 70 tCO2e. A number of things aren’t factored into these calculations: staff with green energy tariffs, staff with solar panels and home batteries, or other renewable energy sources, and the different needs for heating and air conditioning in different countries.
We decided not to include these figures in our overall emissions for 2024, but we are looking at a more reliable way to estimate this for 2025. However, we need to consider what we would do with the information and whether we would, or could, do anything to reduce this.
Hosting services
We use AWS for hosting our REST APIs, Metadata Search, and the website. In 2024, our main metadata registry was in a data centre in Massachusetts, which is not included in our calculations. In July 2025, we transitioned fully to AWS, so from 2025 onwards, our emissions from AWS will be higher and will encompass our entire system.
In 2023, Amazon reports Crossref’s carbon emissions were 0.216 tCO2e compared with 0.266 tCO2e in 2022. In 2024, emissions were 0.132 tCO2e.
Compared to travel, the footprint from AWS is minimal.
Online meetings
As a distributed, remote-first organisation, Crossref is a heavy Zoom user––it’s essential for staff and for engaging with our community. However, Zoom doesn’t provide tools or estimates of the carbon impact of Zoom meetings. We used a tool last year to provide an estimate, but we aren’t confident it’s accurate or meaningful. The tool was built in 2020 and made a lot of assumptions and guestimates.
To calculate emissions for flights and train journeys, we chose to use Carbon Calculator. For hotel stays and home working estimates, we used Circular Ecology. For AWS, we used the Customer Carbon Footprint Tool (CCFT) provided by AWS.
Offsetting
We don’t offset our emissions from travel or other operations and don’t have plans to do this. Offsetting emissions is problematic in a number of different ways, so we don’t feel confident in doing it.
In conclusion
In general, it feels good to have had a few years of tracking this, learning more, finding the right tools, and trying to stick to a target to limit our increases. While of course there are always reasons for the target to increase—as we grow and are able to subsidise others beyond our staff more—we remain committed to not just monitoring our carbon spend but also maintaining it at a reasonable level and finding ways to limit and mitigate our impact on the environment. This kind of sustainability isn’t included in the POSI Principles for open scholarly infrastructures, but we’d love to see other similar organisations share their tips and measurements so that, as a community, we can learn how to do even better.
