Climate Data Arks
We must save copies of important climate science data sets.
Anyone who has spent any time in management has heard the maxim "You can't manage what you don't measure" ... or some variation of it. Widely attributed to either Peter Drucker or E. Edward Deming, there's no proof either of them actually said it.
William Thompson, Lord Kelvin, however, in 1883 did have this to say about the importance of measurement and science:
"I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely, in your thoughts, advanced to the stage of science, whatever the matter may be".1
Nowhere is that more true than for the data that underpins climate science. Without credible data, science can not be done — the science that informs public opinion, policy, and underpins the best solutions.
Some of that hard-won data exists in only one place, which makes it vulnerable to loss by accident, war, or political action.
These kinds of threats to important data are not hypothetical. Political ideology has already destroyed climate data, even in democracies.
In Canada, under former Prime Minister Stephen Harper, seven out of eleven Department of Fisheries and Oceans marine libraries were closed between 2012 - 2015, with the majority of their archived materials being destroyed or discarded without being digitized. This was part of a larger effort to muzzle government science and scientists on topics related to the environment and climate change.2
Nearly a year ago, we wrote about the current US administration's actions to eliminate scientific (and other) data sources that don't align with its beliefs, and about some of the ad-hoc efforts to copy that data and save it somewhere other than within the US government.
Now comes word of a project outside of the USA to copy and archive climate simulation data.
As reported in a recent story3 in the online magazine Yale Environment 360, published by the Yale School of Environment, the German Climate Computing Center (DKRZ) has launched a program called "CMIP7-ark", specifically to save and store backup versions of climate simulations done by US scientists in support of the seventh major United Nations IPCC climate assessment, due in 2029. According to the article, Europeans are worried that the Trump administration may attempt to block, delete, or skew the data. Other projects are underway to also back up other US science data sets.
The challenges of moving the data are significant. Some data sets are many petabytes in size, exceeding the capacity of even high speed digital connections. Copying to magnetic tape or hard drives and physically carrying them by hand may be the most practical method.
The Yale Environment 360 article notes the problem of potential data loss is not confined to ideological manipulation in the USA, citing the example of how the war in Ukraine threatened a valuable baseline set of 19th century meteorological observations in that country.
The article also quotes the president of the American Geophysical Union, who notes these kinds of efforts amount to a form of triage, and the science community must now fundamentally reimagine data organisation and maintenance.
One potential threat is not addressed in the article. The head of the CMIP7-ark project says that in an ideal world, science data would be connected by a digital pipeline for easy backup. This may not be wise in a world of fast-evolving AI models who are escaping their guardrails. It is easy to imagine one of these models accidentally, or intentionally when prompted by a bad actor, gaining access to such interconnected data and altering or deleting it.
It may be time to begin giving science data the same protections we afford reserves of gold bullion. In the end it may be more valuable.
Reading
- Stellman, Jeanne Mager (ed.), Encyclopaedia of Occupational Health and Safety, 4th edition, Vol. II, Geneva, International Labour Office, 1998
- Robertson, Manjulika E., Bringing Evidence Back from the Dead: A History of Interference in Science in Canada, Dalhousie Journal of Interdisciplinary Management, Volume 17, 2023. ojs.library.dal.ca/djim
- "Why Germany Is Building an Ark for U.S. Climate Data," Yale E360, accessed September 26, 2026, https://e360.yale.edu/features/germany-us-climate-data.