PurPassionSky

Climate Data Rescue Scientist / Historical Records Archivist

Unexpected

The pull to make things last

Preservation · Ephemeral → Enduring

The shape of the work

Pace
  • Patient work over a long time, where showing up matters most
  • A steady rhythm with room to breathe
  • Short, intense, and the stakes are right now
What your week looks likeQuiet stretches, then deadline storms
How much you move around at workMostly at a desk, but you get up for site visits or other work
Whether you can work from anywhereSome days in, some days wherever you want
How quickly you receive feedback on your workYou might wait years to see if it mattered
What you're actually working withNumbers, measurements, records — things you read on a screen / Concepts, theories, designs, stories — things you think up / Materials, organisms, land, equipment — things you can touch

Behaviours this work is made of

Core
  • Creating lasting records.
  • Creating the record that allows others to understand later.
  • Maintaining the state of something over time.
  • Capturing moments, sounds, events.
Also present
  • Working with others toward a shared thing.
  • Systematic, methodical pursuit of understanding.
  • Creating consistency at scale.

In words

Before satellites and automated sensors, somebody wrote the weather down by hand — a ship's officer logging temperature and wind three times a day in the 1880s, a colonial weather station keeping a paper ledger, a lighthouse keeper recording pressure. Millions of these records exist, in archives, libraries and boxes in basements, and almost none of them are in a form a climate model can use. This role finds them, works out whether they're worth recovering, and turns handwriting on decaying paper into a dataset alongside modern instrumental records.

The distinct contribution is length. The Palaeoclimatologist in this field (see archetype 1) reads climate from physical archives like ice and sediment that go back hundreds of thousands of years but can't say what a specific July was like; the instrumental record most climate science relies on starts properly only in the mid-twentieth century. Data rescue extends the reliable instrumental record backward by decades, sometimes centuries, which matters directly: a model tested against a longer record is a better model, and an extreme event only looks unprecedented if you know what happened before anyone was measuring properly. The WeatherRescue.org project digitised more than three million daily weather observations from the Scottish Highlands between 1883 and 1910; in 2026 the Royal Society digitised its own historical meteorological archive, and a separate archive of more than 1,600 sets of meteorological, magnetic and tidal observations spanning 1706 to 1915 was made available for the first time [from salary and job-posting data, WeatherRescue.org / Royal Society, 2026].

The work increasingly pairs archival skill with technology built for exactly this problem. Machine-learning transcription tools such as MeteoSaver are now used to read historical weather tables automatically, and citizen-science projects like ReData recruit volunteers to transcribe digitised bulletins at a scale no institution could staff alone — the archivist's job has shifted toward managing and quality-checking that pipeline as much as doing the transcription by hand [MeteoSaver / ReData, 2026].

There's a guide here if you want one

Kitsune can talk through anything on this page — whether it might suit you, what to do next, questions this page doesn't answer. Everything here is yours to read either way.

The expertise is scarce in a way that should worry the field more than it currently does. Recovering a historical archive takes the skills of a historian, an archivist and a scientist at once — knowing where records might exist, reading old handwriting and obsolete units, and judging whether the data is trustworthy enough to use. The people who can do all three are few, and the number is shrinking rather than growing [WMO, 2026].

The funding case is genuinely hard to make. Data rescue produces no new finding of its own — it produces the raw material other people's findings will later depend on — which means it competes badly for grant money against research that promises a result. Archives are lost to funding gaps as often as to flood or fire.

And some of it is a genuine race. Paper degrades, buildings get cleared, and an archive nobody has catalogued can be discarded by someone who has no idea what it is. A record that survived a hundred and fifty years can still be lost in an afternoon by someone doing an unrelated clear-out, which is why this role is as much about finding archives before that happens as it is about the transcription itself.

A degree in a numerate physical or environmental science (physics, geography, atmospheric or environmental science) is the usual base, sometimes paired with archival or library science training or a genuine interest in historical research. Programming and data-handling skills matter for building and cleaning the resulting datasets. Entry routes include national meteorological services, national archives and libraries with historical science collections, and international efforts such as the Copernicus Climate Change Service's Data Rescue Service, which coordinates recovery projects and increasingly works with citizen-science platforms [C3S Data Rescue Service, 2026]. Because so much of the current effort runs on volunteer transcription, getting involved in an active citizen-science data-rescue project is a genuine and accessible way to find out whether the work suits you before committing to it as a career.

Before you go