New AI-Based Analysis Traces Variscite's Journey Through Ancient Europe

For thousands of years before diamonds or emeralds dazzled Europe, one of the most prized gemstones was a vivid green mineral called variscite. Ancient communities treasured it for its luminous color and carved it into beads, pendants, bracelets and rings. But while archaeologists have long uncovered variscite jewelry across Western Europe, one question remained stubbornly difficult to answer: Where did these eye-catching green stones actually come from, and how did they travel so far from their points of origin?

A team of archaeologists and artificial-intelligence specialists from Spain and Portugal has now brought modern technology to this ancient mystery. Their newly published study combines traditional fieldwork with advanced chemical analysis and machine-learning tools to trace variscite artifacts back to their geological sources. The result is a clearer picture of prehistoric trade networks — and an unexpected story of how these green gems moved across Europe thousands of years ago.

Variscite, whose vibrant green color can range from pale mint and seafoam to deeper emerald hues, is naturally rare and was especially valued between the sixth and second millennia BC. Its color made it a symbol of status and personal adornment. Until recently, researchers could only make educated guesses about where a particular bead or pendant had originated. Mines in Encinasola, Gavà, and Aliste on the Iberian Peninsula were all considered likely sources, but the relative importance of each site was unclear.

The new research changes that. The team began by gathering more than 1,800 geological samples from known variscite deposits and comparing them with 571 artifacts recovered from archaeological sites across modern-day Spain and beyond. Each sample was analyzed using portable X-ray fluorescence, a non-destructive technique that measures the tiny chemical variations locked inside the stone.

These chemical fingerprints — slight differences in elements such as aluminum, phosphorus, and trace minerals — turn out to be unique enough to identify the mine of origin. But because the dataset was so large, the researchers used a machine-learning method called a random forest algorithm to find patterns no human could easily spot.

In simple terms, the AI learns the “signature” of each mine and then predicts where an ancient bead came from. The system achieved 95% accuracy, a remarkable level of precision for artifacts thousands of years old.

With the origins finally mapped, a new trade picture emerged. The study shows that the Gavà and Aliste mines were far more important than previously believed, while Encinasola played a smaller role. Even more surprising, variscite beads found in northern France likely traveled overland from northern Spain, crossing the Pyrenees along long-distance routes. Earlier theories had emphasized maritime trade, but the AI-driven analysis suggests that land-based exchange networks were significant and highly organized.

The project’s open-science framework — making all data and code freely available — sets the stage for future work. Researchers hope to apply the same approach to other materials, from amber to early metals, building a richer understanding of how beauty, craftsmanship and commerce shaped prehistoric societies.

And while the ancestors who wore variscite couldn’t have imagined artificial intelligence, their cherished green gemstones are now helping scientists tell a very modern story: how technology can illuminate the human connections of the distant past.

Credits: Photo of Smithsonian specimen by Jstuby at en.wikipedia, Public domain, via Wikimedia Commons. Photo of gold and variscite 2nd century AD earrings by Metropolitan Museum of Art, CC0, via Wikimedia Commons.