A web tool that puts every fossil back on the map it actually came from

A fossil dug out of a roadside cutting in Scotland can sit on land that, 400 million years ago, was a tropical lagoon south of the equator. A palaeontologist holding that specimen in a museum drawer is holding two places at once: the spot where it was found, and the spot where the animal actually lived. For most of the discipline's history, reconstructing that second place required specialist software, a working knowledge of plate-tectonic reconstructions, and a great deal of patience. As of mid-July 2026, it requires a web browser.

On 20 July 2026, researchers released an open-access tool that lets a user drop a present-day latitude and longitude into a form, pick a geological time slice, and watch the point rotate backwards through deep time onto the continent where the rock originally accumulated. The output is a map, a set of coordinates, and a downloadable reconstruction suitable for plotting alongside fossil occurrences in a paper. The release matters less for any single result than for what it removes: a tedious, error-prone chore that quietly shaped which questions got asked and which got shelved.

What the tool actually does

At its core, the application is a front end on top of a long-running body of work: plate-tectonic reconstructions that describe, in roughly 10-million-year increments from the present back to the early Palaeozoic, where each modern landmass was positioned relative to the others. The user enters a modern coordinate, a formation age, and a rotation model. The tool then applies the appropriate series of rotations, in reverse, to send that point back to its position on the parent continent at the moment the sediment was laid down. It is, in effect, the inverse of the workflow that has long existed for geophysicists plotting present-day earthquakes onto ancient terranes, simplified into a form that a working biologist can use without writing code.

The publication reporting the release frames the practical payoff in straightforward terms: determining where a fossil was located in the geological past is a computational challenge that has long demanded refined numerical methods, and the new tool makes the process accessible in seconds rather than weeks.

Why the inverse problem has always been hard

Plate-tectonic reconstructions come in families, and the families disagree. A point rotated back under one widely used model can land on the eastern margin of a vanished microcontinent; the same point under a rival model lands on the western margin of the same landmass, or on a different landmass entirely. The choice of model is therefore not a neutral preference. It is a hypothesis about Earth's history, and it can move a fossil from one biogeographic province to another.

This is where the bottleneck has historically sat. Running a coordinate backwards through a reconstruction by hand meant consulting tables of rotation parameters, applying Euler poles in sequence, and double-checking the arithmetic. The arithmetic, in turn, was a known source of small errors that propagated into biogeographic arguments about whether two faunas were neighbours or whether a species dispersed across an ocean. Tools existed, but they lived on individual researchers' laptops, in idiosyncratic formats, and required fluency in the underlying plate model.

What changes when the chore disappears

The structural effect of moving a task from specialist software to a web form is familiar from other fields. Search engines did not invent the question "what is the capital of Burkina Faso"; they collapsed the cost of asking it. Genome browsers did not invent comparative genomics; they collapsed the cost of looking. The fossil-rotation tool is the same kind of move at a much smaller scale.

For palaeontologists working on biogeography, the immediate gain is coverage. Studies that previously had to pick a handful of localities to reconstruct, because each one cost an afternoon, can now reconstruct hundreds and look for statistical patterns in the resulting dataset. For researchers working on the deep-rooted question of how terrestrial ecosystems assembled as continents collided and split, the gain is comparability across studies: a common tool with documented models means a result from one lab can be checked against another without having to re-derive the rotation chain.

For the wider public-facing side of the discipline, the tool also changes what museum and university communications teams can do. A specimen label that once read "found in Moray" can now plausibly read "found in Moray; lived near the equator, 410 million years ago", with the second half of the sentence derived rather than asserted.

The caveats the tool does not remove

The release does not resolve the underlying disagreement between competing plate models, and it is not intended to. A user still has to choose which reconstruction to trust, and the choice carries scientific weight. Nor does the tool address the question of how confident one can be in the age assigned to a particular fossil locality: a rotation is only as good as the date it is anchored to, and that date comes from biostratigraphy, radiometric dating, or both. If the formation age is wrong by a few tens of millions of years, the rotated point will land in the wrong ocean.

There is also a quieter structural concern. The model underlying any such tool encodes assumptions about how continents behaved, and those assumptions are themselves the product of a particular research community. As with any piece of scientific plumbing, the question of who maintains it, who funds the maintenance, and whose plate models get incorporated and whose do not, will shape what kind of palaeogeography the broader field ends up practising. Open release is not the same thing as open governance, and the difference tends to matter over decades.

What is plainly true is that the cost of asking a basic question in historical biogeography has fallen sharply, and that the fall will show up first in the volume of locality-level reconstructions appearing in the literature. Whether the resulting maps describe the deep past more accurately will depend on choices that no browser interface can make on the user's behalf.

This article was produced by Monexus's science desk. Where mainstream coverage tends to frame new digital tools as inevitable productivity gains, this publication looks at the underlying scientific choices the tool cannot make for its users, and at who benefits when a previously specialist chore becomes a browser field.