Purging Scientific Datasets to Fight “Wokeness” and “Waste”
Astrophysicist Clara Sousa-Silva needs data on Earth’s climate to accurately observe space. Earlier this month, she discovered that crucial climate datasets had disappeared. When DOGE cuts accelerated, more data vanished.
A meteor streaks across the night sky at the InfoAge Space Exploration Center in New Jersey on April 22, 2022. (Tayfun Coskun / Anadolu Agency via Getty Images)
- Interview by
- Meagan Day
In early February, molecular astrophysicist Clara Sousa-Silva discovered something alarming: crucial scientific data she needed for her research had vanished from federal government servers. As a scientist who studies the atmospheres of other planets and moons using ground-based telescopes, Sousa-Silva relies on climate monitoring data from the National Oceanic and Atmospheric Administration (NOAA) to filter out Earth’s atmospheric effects from her observations. What began as a personal research obstacle quickly revealed a much larger crisis affecting scientists across disciplines.
This data disappearance coincides with the aggressive implementation of two intertwined initiatives from the Trump administration: the Department of Government Efficiency (DOGE), led by Elon Musk, and President Donald Trump’s campaign to purge the federal government of anything “woke,” including efforts to combat climate change. DOGE has put a target on NOAA’s back, along with other agencies that gather and manage crucial scientific data like the Environmental Protection Agency (EPA). Meanwhile, Trump’s executive orders have dismantled numerous climate initiatives, lumping them in with his increasingly nebulous and sprawling definition of wokeness.
The consequences have been swift and severe. Funding for research mentioning climate change has been withdrawn or modified, while scientists in federal agencies have been reassigned or removed. Climate data and information have been systematically removed from government websites, while access to scientific databases has become increasingly restricted — all under the banner of combating “wokeness” and improving “government efficiency.”
In this interview, Sousa-Silva, a physics professor at Bard College specializing in alien atmospheres, details the breakdown in federally maintained databases in the weeks following Trump’s inauguration. The situation exposes two troubling developments: the targeted removal of climate science information and broader infrastructure damage caused by staffing disruptions at federal agencies. As Sousa-Silva explains, the consequences extend far beyond her research on other planets, potentially hindering weather prediction capabilities and erasing decades of irreplaceable scientific records.
Can you start by explaining the nature of your research?
I study the atmospheres of other planets and moons, mostly by looking through our own atmosphere with ground-based telescopes.
How is your work connected to climate science?
Because I use mainly ground-based telescopes, if I want to look at data that belongs to alien planets in our solar system or beyond, I need to be able to subtract the effect of Earth’s atmosphere. So I need to know it pretty well.
Could you walk us through how you first discovered the data deletions and what specifically went missing?
A couple of weeks ago, I went to look at climate monitoring data that’s collected from ground-based monitoring stations, including one on Mauna Loa in Hawaii. Mauna Loa is the mountain next to Mauna Kea where my telescopes are. We use the Mauna Loa data, which is NOAA Climate Data Record program data, to obtain information about what the sky looks like so that we can remove it from our own data.
But all the data was gone. So I contacted my colleagues who are climate scientists, who basically do the same work as I do, but instead of doing it on atmospheres beyond Earth, they do it for Earth. They started panicking because lots of data was missing, specifically all the CO₂ data was missing.
We then learned that there had been some scheduled maintenance of some databases, but even after the databases came back online, there was still stuff missing intermittently. At one point, a lot of the greenhouse gases data was missing. Then they came back, but not CO₂. Then CO₂ came back.
I contacted my colleagues who use Environmental Protection Agency (EPA) data, and they reported the same thing — their repositories were not accessible. I found out from my colleagues who work at NOAA that there hadn’t been a centralized effort to safeguard this data, which means the only access people have is through online databases. If those become disconnected, even if the data is not deleted, it basically becomes useless.
Did you know immediately when you saw the missing data that this had something to do with the change in presidential administration?
I made an educated guess. But I only really knew once I saw that there had been a DOGE presence at NOAA. I know from my experience in this sort of work that it’s often one or two employees whom no one really appreciates who are maintaining these databases. If they get removed for whatever reason, it’s not necessarily the case that these databases will just go on. They’ll stop working, and no one will know how to bring them back up.
I knew a lot of people were potentially losing their jobs or being suspended or disincentivized to stay. Access to this data would necessarily become precarious. So I started trying to save it. It wasn’t until later that I realized that it wasn’t just data on climate.
A lot of my data from space — Venus and other stars in our galactic neighborhood — which I really thought would not be targeted in any way, that data started disappearing. A colleague’s data on Jupiter started disappearing. My only presumption is that it’s federally funded, so disruptions in federal agencies are causing the disappearances.
The taxpayers own that data. I’m hoping no one’s deleting the data, but it doesn’t really matter if you don’t delete the data if you make it inaccessible. It’s almost as if you did, because no one will remember any of the file names or the directory structures to get them back.
It sounds like there are two layers to this issue: politically motivated attacks on climate science data and also broader disruption to federal workers and infrastructure through DOGE, which has caused data maintenance issues. Is that correct?
Exactly. At the very least, this data is in jeopardy. To be fair, it has been intermittently coming online and offline in the past few weeks. I’ve tried to save as much as I can myself, but I can’t even fully use the data because some of it needs experts to direct you to the scripts needed to read it, and those experts are mostly unresponsive.
The staffing issue has consequences for access to things that you would think are passive — meaning the data’s already been obtained, it’s been paid for, it just sits in a repository. But actually, the maintenance of the accessibility of the data requires people. If people are being laid off or made to prioritize other work, then it’s as if the data never existed.
What are you and your colleagues doing to address this issue?
I’m working to save as much data as possible, as it comes back intermittently. But I wish there were a centralized effort to make it available some other way. Well, actually, there was — it was the federal government! That’s how it should be.
I don’t have the technical power to make even my backups available to the rest of the world like the federal government used to do. I almost feel like I now have a selfish little pile of climate data that I don’t really know how to distribute without putting a target on my back. I’m hoping I’m not the only one safeguarding all historical CO₂ data above American monitoring stations. Surely someone else is too, but that’s not clear to me.
How badly will this impact your research if it disappears, and what are the ripple effects throughout the scientific community?
The consequences for my research on Venus are pretty bad. I will be a worse astrophysicist. But that feels pretty minor considering the importance of being able to monitor how our climate is changing. To climate science, this is absolutely disastrous.
These molecules are not just important to track climate change. They’re important to predict the weather. Even if you don’t think that climate science is a worthwhile effort, I’m fairly certain we can all agree on the bipartisan issue that it’s good to be able to predict the weather. That’s how bad this is. If we lose access to this data, our ability to predict both common rain and extreme weather events becomes jeopardized forever because the historical record is how we model trends.
It’s very hard not to feel hopeless, because it takes decades to establish a working system to access this data. It takes so long to set up a system like data accessibility, but it’s so quick to bring it down. You just need to fire a few crucial people or disconnect a server, and that’s it. It’s lost.
Do you think that Trump and his administration even know the impact that they’re having on science in the name of “anti-wokeness” and so-called government efficiency?
I think they’re telling themselves, and to some extent telling us, that the plan is to break a lot of things, see who complains, and then fix those things. In principle, I can see how they think that would work. The problem is the staffing needed to break something is one person. The personnel needed to fix something is orders of magnitude greater than that.
Unless they’re planning to hire astrophysicists and climate scientists at one hundred times the number of each of their DOGE destroyers, they can’t fix even the things that they do discover should be fixed. The personnel ratio just isn’t there.