UAS Technology Tackles Rapa Nui
It’s one of the most isolated and intriguing places on the planet.
Sitting alone in the South Pacific, tiny Easter Island (traditionally known as Rapa Nui) has long been the source of fascination — and disagreement — for historians and archaeologists alike. Researchers from around the world have conducted decades of studies to learn about the island’s people, environment and famous statues.
However, an important scientific tool has been missing. In spite of years of research, there is no comprehensive geospatial dataset for Rapa Nui. That is, until a team of California scientists set to work to solve the problem.
Formed by volcanic activity, Rapa Nui lies more than 2,200 miles off of the South American coast. The nearest inhabited land is Pitcairn Island, some 1,300 miles to the west.
Rapa Nui’s original human residents arrived from Polynesia around 1200 AD and the first European explorers landed in 1722. Today, roughly 60 percent of the island’s 5,800 residents are direct descendants of the Polynesian settlers. The island was annexed as a territory of Chile in 1888.
Rapa Nui’s most famous occupants are more than 900 stone statues (known as moai) located across the island. Carved from exposed outcrops of basalt and volcanic tuff, the moai were walked to their present locations, where they sit on massive stone platforms called ahu. On average, the statues stand roughly 13 feet high and weigh 13.8 tons, but some are almost 30 feet high and weigh more than 82 tons. In 1935, the Chilean government established the Rapa Nui National Park covering about 40 percent of the island’s 62 square miles. In 1995, the park was designated as a World Heritage Site by the United Nations Educational, Scientific and Cultural Organization (UNESCO).
The significance of the moai, together with questions of how they were created, moved and placed, attracts tourists and scientists from around the world. For many researchers, understanding the moai requires gaining a bigger picture of the lives of the prehistoric residents of Rapa Nui. According to Carl Lipo, professor of anthropology at California State University Long Beach (CSULB), the entire island is a mystery. “Rapa Nui has very few natural resources,” Lipo says. “There are no streams, poor soils, few native species of birds, and only a tiny reef to provide marine resources. When you look at it from a European perspective, it’s surprising that people lived there for any length of time.”
Learning about the prehistory of the island begins with the land itself. Rapa Nui is dotted with evidence of its prehistoric inhabitants, including ruins of houses, ovens, gardens and cultivation features. While Lipo has conducted significant research on the moai, his current research focuses on freshwater resources and its relations to archaeological settlements. “The topography is the key to understanding the archaeological record,” he explains. “It helps you locate where the water and arable land were located.” Good topographic data, he adds, helps to spot ancient roads and building sites that are undetectable from the ground.
“The entire island is an archaeological resource,” says Suzanne Wechsler, an associate professor in geography at CSULB. “Understanding the features and their spatial relationships requires a systematic landscape-scale survey.”
In most locations around the world, aerial imagery is used to supply the high-resolution topographic information that Wechsler describes. But isolation and limited research budgets have thwarted effective aerial photography on Rapa Nui. Satellite imagery is available, but it doesn’t provide the needed resolution.
In attempts to collect useable imagery, CSULB research teams have experimented with a variety of aerial platforms, including kites, blimps, and hobbyist-grade fixed-wing and quadcopter aircraft. The resulting images could provide visual reference over limited areas, but were not suitable to develop a digital elevation model (DEM) or provide accurate measurement of objects and features.
To obtain higher-quality data, the researchers turned to a commercial unmanned aircraft system (UAS) designed to capture the systematic, georeferenced imagery needed to create accurate, highquality maps and terrain models. Lipo,who has years of experience using UAS for archaeology, was confident that offthe- shelf solutions could obtain data suitable for scientific work.
To prove the effectiveness of UAS technology, the team conducted a project along the southern coast of Rapa Nui in January 2015. With funding from the U.S. National Science Foundation, the objectives were to evaluate the performance of the UAS in capturing aerial imagery and to integrate the resulting orthophotos with existing datasets. They selected a Trimble UX5 unmanned aircraft system to collect imagery along the southern coast.
“In nine days, we flew more than 26 missions covering an area of approximately 18.5 square kilometers (7.1 sq. mi.),” Lipo says. “The UAS captured more than 20,000 individual images that were processed to produce 26 orthophotos.” Flying at an elevation of approximately 330 feet above ground level with 80 percent overlap, the images produced a ground sample distance of 2 to 3 centimeters.
As part of its planning, the team identified a minimum of five ground control points for each flight. The stations were measured with Trimble GeoExplorer 6000 GNSS handheld computers, and post-processed to decimeter accuracy using data from a GPS reference station that was originally established for the U.S. space agency NASA. (In the 1980s, Rapa Nui’s airport runway was expanded to serve as an emergency landing site for NASA’s space shuttle.)
The CSULB team also operated its own GNSS base as a backup, Lipo says. The GNSS stations provided a consistent reference frame for positioning of features and ground control points. There are a number of geodetic reference points on Rapa Nui; by tying into them, the scientists can develop absolute locations for the archaeological and hydrology features.
Working with his CSULB colleague Prof. Christopher Lee, Lipo operated the UAS and completed three or four flights each day. They used the Trimble Access Aerial Imaging application to define polygonal coverage areas for each flight. The polygons allowed them to optimize the flights to cover the near-shore areas, and collect some inland data as well.
The researchers say the planning was very helpful in getting the best coverage in the face of constant winds, irregular coastline and rocky terrain that limited the selection of landing sites. In several cases, they launched multiple missions from the same location, with the aircraft flying a kilometer or more to the target area before beginning photo passes. Between flights, the team downloaded images and installed a fresh battery into the aircraft. In some cases, they switched cameras, replacing the high-resolution color camera with a near-infrared sensor. “It’s a fast change,” Lipo says, “and you can reuse the previous flight plan. So, you can easily match the two flights for lighting and weather.”
Working to manage the large datasets — the system produced 60Gb of imagery — they produced orthophotos from whichthey derived DEMs for topographic analysis. They then used Trimble eCognition software to classify archaeological features. The software uses object-based analysis to identify the remains of houses, stone platforms and circular structures for gardens.
The performance of the Trimble UAS — both in the air and the information it produced — convinced the CSULB team of the value of using commercial solutions to extend the aerial imaging over the entire island. The team is planning new projects that will use the Trimble system to blanket all of Rapa Nui with high-resolution imagery.
In addition to archaeology, aerial images can support other activities on the island. The information can assist the island’s Ministry of Public Works by documenting modern infrastructure and providing information for planning and development to handle the island’s growing business in eco-tourism. For example, the ministry is planning bike paths around the island. It can use the detailed data to design paths to access historic features without damaging the archaeological record. Because the imagery can be repeated at low cost, it enables researchers and government officials to see changes in the features and topography over time to gauge the impact of visitors and development.
For Lipo, the payoff lies in the science. By combining topographic information with hydrological data, he has gained new insights into Rapa Nui’s history. To illustrate, he points out that the sites of the moai and ahu were apparently based on the location of water rather than for visibility as previously believed.
“The data are astounding,” he says. “You see things that you could never see before, even though the island has been studied for hundreds of years. The UAS provides a complete record of what is on the ground. It’s the way archaeology should be done.”