Published: May 5, 1997

A team of University of Colorado at Boulder researchers have combined a unique set of remote-sensing techniques to pinpoint a series of eight previously unknown Paleo-Indian quarries in Montana dating back as far as 10,000 years.

Led by anthropology doctoral student Thomas Carr, the team used satellite images, aerial photographs and ground-based geophysics to develop a new method for locating ancient Indian mining pits from space. The quarries, the largest of which is about 200 meters on a side, were used by ancient North Americans to obtain a quartz-like material known as chert for tool-making.

A paper on the subject by Carr and research associate Mort Turner of CU-Boulder’s Institute of Arctic and Alpine Research was presented at a recent conference at the University of Pennsylvania Museum of Archaeology and Anthropology in Philadelphia. The field research team also included four CU-Boulder undergraduate and graduate students.

The researchers used aerial photographs, geologic maps and ground tests to develop “spectral classes” characterizing 12 distinct geologic and vegetation types found in the Horse Prairie Valley study area in southwest Montana. Their data included electromagnetic measurements of soil conductivity from a known Indian quarry in the area called the South Everson Creek site, which is characterized by “altered bedrock” containing veins of chert and an adjacent area of soil disturbed by quarry activity, said Carr.

The team ran the data on a computer software package known as ENVI developed jointly by Research Systems Inc. of Boulder and CU’s Center for the Study of Earth from Space. The software compared nearly 1 million 30-meter-bymeter pixels, or picture elements, from a 1985 Landsat satellite image of the area to each of the 12 spectral classes, identifying the most likely matches.

The researchers speculated that pixels of altered bedrock located next to pixels of disturbed soil like those at South Everson Creek would be the most likely places to find ancient quarries. Out of 12 possible sites identified by the team in the 800-square-mile research area, eight were confirmed to be ancient Indian quarries during a subsequent ground survey using GPS satellite receivers for accuracy, Carr said.

The four other sites identified as possible Indian quarries turned out to be naturally eroding ridges containing large amounts of opal-based silicates similar to the chalcedony-based chert from Everson Creek. “The four failures actually provided the strongest evidence for why our technique worked,” said Carr.

Although the spectral classes developed by the CU team are specific to Montana’s Horse Prairie Valley, researchers could conceivably use the technology in many other regions of the world to pinpoint ancient quarrying activity, he said.

While Landsat satellites “see” in six spectral bands -- three in visible light bands and three in the infrared -- next-generation satellites will see in more than 200 spectral bands, making future searches “much more sophisticated,” said Carr.

As part of his doctoral thesis, Carr plans to use a portable spectrometer to analyze the chemical signatures of individual artifacts from the Horse Prairie Valley region, which may allow him to track known artifacts back to specific quarries.

“It looks like the heaviest use of these particular quarries came during the Paleo-Indian and Archaic Periods, from about 10,000 years ago to about 5,000 years ago,” Carr said. “My suspicion is that the South Everson Creek quarry was heavily used during that period, and an overflow population during the Archaic period caused some of the people to begin working the outlying quarries.”

Indian artifacts made from South Everson Creek chert have turned up 200 to 300 miles away, including the Snake River Valley of Idaho, the Columbia River Basin and the Northern Great Plains. The project should shed new light on Native American settlement, trading and migration patterns in the region, he said.