Angkor Underground: GPR Analysis in Cambodian Archaeology

A review of Angkor Underground: Applying GPR to Analyse the Diachronic Structure of a Great Urban Complex, by Till F. Sonnemann.

Although most archaeologists are familiar with the ancient city of Angkor, they mostly only think of the surficial archaeological record of the site, not the subsurface archaeology. Because Angkor is one of the largest continuous archaeological landscapes in the world and was inhabited for hundreds of years, understanding the subsurface archaeological record is a challenge for archaeologists. Till F. Sonnemann in his dissertation Angkor Underground: Applying GRP to Analyse the Diachronic Structure of a Great Urban Complex presents a cohesive argument for the use of geophysical methods, mainly ground penetrating radar (GPR), to efficiently explore the subsurface archaeological record of Angkor. Walking over 250km of Angkor’s landscape, Sonnemann’s GPR survey covered over 12Ha to test Bernard-Philippe Groslier’s  “hydraulic city” hypotheses and further articulate the Angkor’s subsurface archaeological record. In short, this dissertation outlines an effective method to explore the sprawling archaeological site of Angkor.

This dissertation consists of an introduction, eight body chapters, a conclusion and one appendix. The main body of the dissertation can roughly be divided into three sections: 1) the first section (Chapters 1 through 4) are background chapters on the geological and cultural history of Angkor; 2) the second section (Chapters 5 and 6) focuses on the application and integration of GPR with satellite remote sensing technologies; 3) the third section (Chapters 7 and 7) focuses on discussing the data collected at Angkor during several seasons of fieldwork.

In Chapter 1, The Angkor Region, Sonnemann places his research within a broader geological, cultural, and archaeological context. Basic geological and geomorphological knowledge of an area is helpful when considering any archaeological site and essential when discussing the application of any geophysical method. His discussion of how soil conditions can influence the results of a GPR survey is of great importance to this type of study. Following this, his review of the cultural history of the area emphasizes that the Angkor landscape is a palimpsest of thousands of years of human activity. With the geological and culture historical context outlined, Sonnemann focuses on the current state of archaeological research at Angkor and ongoing debates around the site such as: 1) how the water management network and the baray system functioned; 2) outlining the boundaries of Angkor; 3) deconstructing the palimpsest of an urban center.

The focus of Chapter 2, Changing Perceptions of Angkor, is the history of primary documentation and archaeological research at Angkor. Starting with the impressions of some of the first visitors to the site, such as Zhou Daguan, a Chinese diplomat who visited the site from A.D. 1296-1297, Sonnemann explores how people’s perceptions of the Angkor landscape have changed through time, and, of particular interest, how archaeological data have modified these perceptions. Bernard-Philippe Groslier’s groundbreaking work greatly changed the way archaeologists and historians viewed the site of Angkor. Due to Cambodia’s political circumstances in the 1970’s and 1980’s, archaeological research halted and much of Groslier’s original data was lost. Excavations resumed in the 1990’s and these data have contributed to discussions of low-density urbanism and human environmental modification. Figures 15 (pp. 69) and 16 (pp. 70) illustrate the development of archaeological research and changing perceptions of Angkor. Sonnemann’s discussion of the history of Angkor clarifies the history of exploration at Angkor and places his research within a historical context.

In Chapter 3, GPR, Processes and Procedures, Sonnemann introduces the methodological and technical aspects of conducting a GPR survey at Angkor. The chapter is divided into two sections: 1) Sensing a Landscape Beneath the Surface; 2) Application of the Technique. In this first section, Sonnemann reviews the relevant literature regarding to geophysical methods, with an emphasis on GPR. The second section focuses on the specific challenges of using GPR throughout Angkor by exploring three known subsurface contexts: habitation areas, production sites, and burials. At each location, Sonnemann deploys GPR to document the anomalies produced by these different human activities to help identified anomalies from unknown subsurface archaeological contexts. His results suggest that production areas, namely smelting sites, are easiest to detect using GPR. Detecting habitation areas depends largely on the type of construction material that was used in the past. Surprisingly, burials often went undetected. At burial sites, previous looting was easier to detect than known burial contexts. The second section of chapter three is essential background for future applications of GPR at Angkor.

In Chapter 4, GPR and Remote Sensing, Sonnemann explains how he links the GPR data he acquired into an integrated GIS. Sonnemann discusses the history of mapping at Angkor concluding with the introduction and usage of new satellite imagery, such as TerraSAR-X images. Sonnemann cogently argues that site selection for GPR survey heavily relies on accurate and available archaeological and topographic maps. By integrating the GPR data within a GIS, anomalies are more easily classified and can be interpreted with their spatial context.

In Chapter 5, Shrines and Temples, Sonnemann discusses the construction methods of shrines and temples around Angkor to explain the parameters for finding these features in his subsurface GPR survey. Almost all surficial archaeological remains at Angkor are either shrines or temples, yet, the subsurface record of these features remains unclear. Sonnemann targets the western area of Angkor to search for subsurface masonry structures. Using the grid subsurface survey method, he found the foundations of six tower bases around the area of Gopura 4 West. This discovery alters the architectural history of the site and again emphasizes the importance of imagining Angkor as a palimpsest landscape.

In Chapter 6, Enclosures in the Periphery, Sonnemann expands his GPR survey area to include the enclosures of three peripheral areas, the hilltop temple of Chau Srei Vibol, the mounds of Banteay Sra, and Prasat Komnap. Since the surficial archaeological record of the enclosures (the area in between the entry gates and the central temple) has little topographical variation representing masonry architecture, archaeologists know little about these areas. For this reason, Sonnemann conducted GRP survey in cleared areas within the three enclosures. He found a variety of subsurface archaeological remains, such as roads, building foundations, and the remnants of canals.

In Chapter 7, Reservoirs and Infrastructure, Sonnemann addresses the hydraulic city hypothesis using insights from his GPR landscape survey. Some archaeologists remain skeptical that the large barays at Angkor functioned as a water management system. Instead, they believe the barays were primarily had a religious function because they lack evidence of distribution features, such as sluice gates. Sonnemann’s 2-dimensional GPR landscape survey focused on the four major barays at Angkor. Sonnemann’s subsurface data does not show any sluice gates, however, but only one small masonry canal in the East Baray where the existence of a sluice may have been a possibility. He suggests that the water manangement of the baray would have provided a stable water table and could still have been used to a certain amount of irrigation without sluices. Sonnemann’s research contributes to mounting evidence that supports the functional role of the barays in redistributing water for utilitarian purposes.

In Chapter 8, Landscape and Network, Sonnemann aims to further articulate the network of canals and embankments identified at Angkor through surficial remote sensing methods. He combines the GPR dataset he generated with TerraSAR-X imagery to demonstrate that the canals and embankments identified from topographic differences have a clear subsurface component to them. The anomalies that he records clearly differentiate natural streambeds from artificial canals, another important step forward in understanding how the canal system at Angkor operated.

In conclusion, Sonnemann’s dissertation sheds light on large parts of the subsurface archaeological record of Angkor. In the past, research has almost exclusively focused on advances in mapping and recording surficial architecture. Sonnemann’s dissertation Angkor Underground: Applying GPR to Analyse the Diachronic Structure of a Great Urban Complex is a compelling case study for why and how GPR can be used to articulate the subsurface archaeological record of an immense urban area. Covering over 12 ha, Sonnemann’s GPR survey revealed previously unknown architectural components, water management features, and other less obvious buried archaeological deposits. Conducting excavations to make these discoveries over such a large area would have been far too costly in time and money. In sum, Sonnemann’s dissertation provides a model for future GPR analyses of archaeological urban areas and helps shift the archaeological focus at Angkor from the surface to the subsurface.

Michael J. Storozum
Department of Anthropology
Washington University in Saint Louis

Primary Sources
GPR data collected at Angkor, TerraSAR-X imagery, published archaeological reports

Dissertation Information
University of Sydney, New South Wales, Australia. 2011. 352pp. Primary Advisor: Roland Fletcher.

Image: GPR cart in front of southern gate of Banteay Kdei temple, Siem Reap Province, Cambodia; photo taken by author.

Leave a Reply