Art manager

Computational predictive model useful for identifying ancient hunter-gatherer sites

Researchers seeking to identify some of archaeology’s most difficult “finds”, including sites used by nomadic hunter-gatherer communities, are harnessing technology to facilitate research.

Archaeologists at the Max Planck Institute and Simon Fraser University (SFU) are gaining new knowledge through a computer predictive model that can assess the likelihood that landscapes contain such sought-after sites. Their work has just been published in the magazine PLOS One.

“Preserving archaeological sites from destruction ensures that history is not lost and is particularly important for communities that have not used written records; but before sites can be protected and studied, they must first be found,” says Rob Rondeau, SFU PhD student in archeology and acting director of the SFU Museum of Archeology and Ethnology.

Chris Carleton, now of the Max Planck Institute for Chemical Ecology, developed the Locally Adaptive Model of Archaeological Potential (LAMAP) – a predictive model that considers the distributions of the values ​​of landscape variables around known sites.

LAMAP uses information from landscape data to estimate the archaeological potential of lands that have not been examined by archaeologists. Carleton has successfully demonstrated the method at large permanently occupied sites in Belize and Turkey.

Rondeau, an experienced underwater archaeologist, connected to Carleton at the suggestion of SFU archeology professor Mark Collard, who supervised Carleton while he was an SFU doctoral student and postdoctoral fellow.

Rondeau’s interest is in finding ancient sites on submerged landscapes off the coast of British Columbia, those occupied when sea levels were much lower than they are today. Their collaboration will play a major role in helping him identify underwater camps of hunter-gatherers from the distant past.

Carleton and Rondeau tested LAMAP in Alaska’s Tanana Valley, with its rich record of hunter-gatherer sites dating back to the last ice age 14,500 years ago. They “trained” the model with landscape data around 90 known sites, randomly selected from the Alaska State Heritage Database.

The pair predicted which parts of the landscape had the highest potential for producing more sites, then returned to the database to assess their predictions. Areas identified by LAMAP as high potential were found to contain many remnant sites in the database, confirming that LAMAP was able to predict preferred camping areas that may have only been occupied by hunter-gatherers for a few days or a few weeks.

Significant computing power was required to undertake the systematic comparison of millions of data points from the 7,000 square kilometer study area. Rondeau traveled to Alaska in 2019 to experience the landscape and will then apply what was learned to the deep waters off the coast of British Columbia.

– This press release was originally published on the Simon Fraser University website