Some 400 km south of Lima, Peru, on the barren Nazca Pampa, lie hundreds (maybe thousands) of ancient geoglyphs.
Considered among archaeology’s greatest enigmas, the Nazca Lines were discovered by chance in the mid-1920s. It took almost a century to discover some 430 of these geoglyphs. And then, in a span of under six months, artificial intelligence (AI) fuelled a spurt of 303 fresh discoveries.
Here we look at the story of the Nazca Lines to see how AI might be ushering in a revolution in the field of archaeology.
Geoglyphs are motifs created on the ground by manipulating surface stones, soil, or gravel. The ones in Nazca — the oldest among which are more than 2,000 years old — were built by removing rocks and earth to create “negative” images. With the topsoil in the desert oxidised to a deep rust colour, the furrows stand in contrast to the ground around them. And since the area receives next to no rain, and little wind, the designs have largely escaped erosion and survived for thousands of years.
The Nazca Lines were discovered by hikers in the mid 1920s, with Peruvian archaeologist Toribio Mejia Xesspe becoming the first to systematically study the Nazca Lines in 1926. Then in the 1930s, as aviation took off, pilots flying over Peru started discovering more geoglyphs which came in all kinds of shapes and sizes, from trapezoids, spirals, and zigzags to stylised figures of hummingbirds and spiders.
Identifying these geoglyphs has always been challenging. For one, they are next to impossible to identify from the ground level. Even from the skies, low-flying aircraft are typically needed to see the geoglyphs, but the lower one flies the less the ground that they can cover. This makes discovering new geoglyphs, especially smaller ones, a very arduous task.
A survey published in September in the journal the Proceedings of the National Academy of Sciences employed a revolutionary new mechanism to find 303 previously uncharted geoglyphs in under six months. The researchers covered more than 600 sq km of area in the Nazca Pampa using low-flying drones, and then deployed AI to analyse their data.
It is not that these images could not have been detected manually, but pilots would have to know exactly where to look. The pampa (South American flatland) is so immense that “finding the needle in the haystack becomes practically impossible without the help of automation,” Marcus Freitag, an IBM physicist who collaborated on the project, told The New York Times.
So how were these new geoglyphs — which are smaller than the ones discovered before — found using AI? A team of Japanese researchers in collaboration with IBM developed an application which can discern the faintest of outlines from aerial photographs. This AI was able to eliminate 98% of the imagery, leaving researchers 2% or 47,410 sites to manually comb through.
The team led by Dr Masato Sakai, an archaeologist at Yamagata University in Japan who has studied the lines for 30 years, eventually narrowed the field down to 1,309 candidates which the archaeologists then visited for “ground-truthing” purposes. Although 303 new Nazca geoglyphs have been confirmed till now, Dr Sakai believes that there are at least another 500 undetected figures.
This spurt of discoveries will provide experts with far greater insights into the Nazca Lines, and the shadow culture which constructed these. Since it did not leave behind the written word, little is known about the pre-Incan civilisation which once prospered in Peru.
Till about the 1970s, archaeologists believed that these geoglyphs had astronomical significance. Some radical theories even involved aliens and ancient astronauts. Archaeologists today, however, emphasise on these geoglyphs’ ritual significance.
Dr Sakai told The New York Times that geoglyphs functioned as sacred spaces for community rituals, and could be considered planned, public architecture. The newly discovered geoglyphs are mainly located along a network of trails that wound through the pampa, and were most likely made to share information about rites and animal husbandry.
But beyond better understanding the Nazca Lines, the latest study also showcases how AI can revolutionise the field of archaeology even in cases where the subject matter is a well-studied and documented UNESCO World Heritage Site.
“The rate of discovery of new figurative Nazca geoglyphs has been historically on the order of 1.5 a year (from 1940s to 2000s). It has accelerated due to the availability of remotely sensed high-resolution imagery to 18.7 a year from 2004 to 2020. Our current work represents another 16-fold acceleration… Thus, AI may be at the brink of ushering in a revolution in archaeological discoveries like the revolution aerial imaging has had on the field,” the study says.
With inputs from The New York Times