A new methodology for geoglyph research ...

Journal of Archaeological Science: Reports 10 (2016) 119–129

Contents lists available at ScienceDirect

Journal of Archaeological Science: Reports journal homepage: www.elsevier.com/locate/jasrep

A new methodology for geoglyph research: Preliminary survey results and practical workflow from the Quilcapampa Geoglyph Survey (Sihuas Valley, Peru) Peter Bikoulis a,⁎, Felipe Gonzalez-Macqueen b, Giles Spence-Morrow a, Willy Yépez Álvarez c, Stefanie Bautista d, Justin Jennings a,c a

Department of Anthropology, University of Toronto, Canada Department of Anthropology, Western University, London, Ontario, Canada Royal Ontario Museum, Canada d Stanford University, United States b c

a r t i c l e

i n f o

Article history: Received 13 April 2016 Received in revised form 2 September 2016 Accepted 5 September 2016 Available online xxxx Keywords: Archaeological Survey and Prospecting Unmanned Aerial Vehicle (UAV) Geoglyphs Ritual Landscape Middle Horizon Peru

a b s t r a c t With the exception of the Nazca Lines, geoglyphs in the Andes have tended to be studied without regard to their position in the landscape. The objective of the Quilcapampa Geoglyph Survey is to better contextualize rock art by identifying and then mapping areas of high concentration of geoglyphs on the broad pampa surrounding the middle Sihuas Valley in southern Peru. This paper outlines our workflow that combines World View 2 satellite images, unmanned aerial vehicle (UAV) photography, and pedestrian survey to rapidly assess a 250 km2 region. Aided by previous surface survey, the satellite imagery effectively located areas of high concentration of geoglyphs that could then be flown over by the UAV whose high resolution camera allows for the capture of features and details not readily identifiable via satellite. The documentation of Sihuas' geoglyphs aids both academic and conservation efforts in this region of Peru. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Despite considerable methodological variation, most rock art research takes individual motifs and scenes as the primary unit of analysis (Bahn, 2010; Whitley, 2011). These approaches, though valuable, inherently pull art from its larger context within the landscape, obscuring potentially illuminating relationships to other geologic and anthropogenic features (Mark and Billo, 2016). In this paper, we present the preliminary results and workflow from the first year of the Quilcapampa Geoglyph Survey in the pampa surrounding the middle Sihuas Valley of southern Peru. The project is designed to quickly assess the regional context of geoglyphs—designs formed on the ground by moving stones or earth—by integrating satellite imagery, unmanned aerial vehicle (UAV) photography, photogrammetry, and pedestrian survey. The methodologies employed in the Quilcapampa Geoglyph Survey provide a model for how extensive areas of rock art can be studied, especially in those contexts where development, erosion, and other factors have put this art at risk.

⁎ Corresponding author at: Department of Anthropology, University of Toronto, 19 Russell St., Toronto, ON M5S 2S2, Canada. E-mail address: [email protected] (P. Bikoulis).

http://dx.doi.org/10.1016/j.jasrep.2016.09.002 2352-409X/© 2016 Elsevier Ltd. All rights reserved.

Many researchers currently use contemporary high resolution satellite imagery (Beck et al., 2007; De Laet et al., 2007; Lasaponara and Masini, 2006; Ciminale et al., 2009; Hadjimitsis et al., 2009; Trier et al., 2009; Parcak, 2007, 2009; Pappu et al., 2010; Contreras and Brodie, 2010; Kouchoukos, 2001; Casana and Panahipour, 2014; Salvi et al., 2011; Sever and Irwin, 2003), data from the historic CORONA missions (Goossens et al., 2006; Fowler and Fowler, 2005; Ur, 2003; Beck et al., 2007; Casana and Cothren, 2008; Philip et al., 2002; Casana et al., 2012; Gheyle et al., 2006) and historic aerial photography (Palmer and Fowler, 2013; Cowley and Stichelbaut, 2012; Bescoby, 2006; Cox, 1992; Barnes, 2003) to map and document archaeological sites. The use of UAVs—known more popularly as “drones”—have only recently become a viable option for researchers with the wider commercial accessibility of units themselves (Ballarin et al., 2015; Swaminathan, 2013; Mancini et al., 2013; Hill et al., 2014; Prins et al., 2014; Smith et al., 2014; Mascort-Albea et al., 2014; Verhoeven et al., 2012; De Reu et al., 2013). The Quilcapampa Geoglyph Survey combines these image sources by first using high resolution (sub-meter) World View 2 scenes to detect easily identifiable geoglyphs and foot trails throughout the study area, and then employing a UAV to capture extremely high resolution images for the purposes of recording and mapping the fuller corpus of anthropogenic features associated with geoglyph concentrations.

120

P. Bikoulis et al. / Journal of Archaeological Science: Reports 10 (2016) 119–129

The geoglyphs of Nazca (Aveni, 1986, 2000; Reinhard, 1985; Silverman and Browne, 1991), which can stretch for kilometers and often clearly visible from space, are by far the most notable examples of geoglyphs in South America They were created within a sociopolitical context of elites orchestrated line buildings that were associated with processionals, feasts, and other ritual events (Conlee, 2003; Vaughn, 2005). The Nazca lines, however, are not the only examples available in the region. The pampa on either side of the Sihuas River also offers up a rich ritual landscape, but one composed of individual elements that are often considerably smaller in scale. Seen from the World View 2 satellite, the mean size of geoglyphs on the Sihuas Valley pampa is roughly 100 m2. The view from the drone and on foot reveals a wealth of even smaller geoglyphs, as well as rock cairns, ceramic smashes, tombs, windbreaks, and other small-scale anthropogenic features. In comparison to either strictly pedestrian or satellite survey, the use of both satellite- and UAVacquired imagery provides an efficient, cost-effective, and often more accurate assessment of Sihuas' highly complex ritual landscape. We report on the results of our first season's efforts to identify geoglyph concentrations via satellite images and then map those concentrations using a UAV. 2. Rock Art of the Sihuas Valley, Peru The Sihuas Valley is located in the Department of Arequipa, in southern Peru (Fig. 1). Limited survey and excavation (Linares Málaga, 1990:

311–354; Linares Delgado, 2009) suggest that the valley was occupied by a small, well-dispersed, and largely egalitarian population of farmers and pastoralists until the Middle Horizon (AD 650–1050). Foreign influence from both the Nasca and Wari cultures entered Sihuas during this period, correlating with rising population, social stratification, and, perhaps, violence (Tung and Del Castillo, 2005; Tung and Owen, 2006). The subsequent Late Intermediate Period (AD 1050–1400) continued these trends, with increasing economic interaction between coastal and highland communities in the department (De la Vera Cruz, 1996; Szykulski, 2010). The most widely known rock art site in Arequipa is Toro Muerto (Linares Málaga, 1999, 2004; Núñez Jiménez, 1986; van Hoek, 2003). Located in the neighboring Majes Valley, the site is composed of thousands of boulders covered in anthropomorphic, zoomorphic, and geomorphic designs that appear to be Nasca- and Wari- influenced (Nieves, 2007). Most scholars suggest the boulders were carved in the Middle Horizon and Late Intermediate Periods by artists who incorporated local and foreign motifs (van Hoek, 2003). Toro Muerto-like motifs are found throughout Arequipa and the themes portrayed in the rock art—dancing, music, death, cosmic balance—suggest that working with stone was one of the means that populations used to navigate through a turbulent era of cultural change (van Hoek, 2010, 2013). This includes the geoglyphs found in the pampa surrounding the middle Sihuas Valley.

Fig. 1. Map of the Sihuas River Valley showing the World View 2 image footprint, the Majes Project Impact Area and major cities.


121

Fig. 2. Map of the Gross-MUNSA.

There are hundreds of geoglyphs on the Sihuas pampa. Most geoglyphs were made by using a negative technique, or “campo barrido”, that creates a figure by removing surface stones to expose the sandy soil below. Other geoglyphs were made by a positive technique in which stones were aligned to form circles and lines (e.g., Linares Málaga, 1990: 320). Eloy Linares Malága has catalogued some of these geoglyphs (Linares Málaga, 1978, 1990, 1999, 2004, 2013), focusing on a few of the more spectacular examples like a 50 meter long winding serpent and a rectangle with stepped frets known as the Gross-MUNSA (G079) (Linares Málaga, 2013: 766–769) (Fig. 2). Most geoglyphs, however, remain unpublished, and there has been nothing written about their placement in the landscape relative to other geological or anthropogenic features. Since 2013, the Proyecto Investigación Arqueológica de Quilcapampa (PIAQ) has studied the rise and fall of Quilcapampa, a site in the Sihuas Valley that grew in regional importance at the end of the Middle Horizon period. In 2014, PIAQ co-director Willy Yépez Álvarez visited the pampa as it was believed that the surrounding geoglyphs might provide insights into Quilcapampa's development. He observed and documented several geoglyphs and noted a possible connection between the geoglyphs and pathways leaving the valley. The Quilcapampa Geoglyph Survey was inaugurated in the same year to further investigate geoglyph concentrations and associations.

3. Outline of work flow 3.1. Preliminary site identification using WORLD VIEW 2 imagery Many archaeological projects now incorporate high resolution satellite imagery in their initial survey design and carrying out field work (Parcak, 2007; Beck et al., 2007; De Laet et al., 2007; Alexakis et al., 2009; Sever and Irwin, 2003). We also turned to high resolution imagery as a first step in our work, obtaining high resolution World View 2 imagery in collaboration with the University of Toronto's Map and Data Library in the fall of 2014. The geoglyphs identified by Yépez Álvarez, once relocated in the acquired imagery, provided a base map for spatially representing and situating these initial sites in their larger geographic context. The satellite imagery was then used to manually identify clusters of likely geoglyphs that could be visited during fieldwork in the summer of 2015.

Our early digital exploratory efforts focused on the mapping of larger, clearly distinguishable geoglyphs that were detectable from direct observation of the World View 2 images. Automated detection processes were not used due to a number of factors, including the fact that the geoglyphs were irregularly shaped which led to an inability to satisfactorily or efficiently differentiate between the imagery digital number (color) associated with ancient and modern features. As discussed further below, we found that manual inspection of the image proved effective in correctly identifying large geoglyphs in the survey area as targets for subsequent reconnaissance and drone image capture. Moreover, later field work would confirm that concentrations of large geoglyphs were reliable proxies for areas of higher anthropogenic activity on the pampa, much of which was not clearly visible in the satellite image but captured later by the UAV. Visual inspection of the World View 2 images revealed 80 geoglyphs in the study region, 27 other architectural features, and 14 unknown features to be inspected in the field (Fig. 3). A preliminary survey strategy was designed following a stratified systematic sampling method to examine each type of feature evenly. This strategy had to be revised in order to adapt to time constraints, geographical distances between features, and especially the technical limitations of the drone unit itself. Instead, the focus was shifted from single features to larger clusters that could be captured completely in a limited number of passes. The identified targets had high spatial autocorrelation,1 suggesting that there was a highly structured distribution of geoglyphs. There is low probability that more geoglyphs than already found on the open pampa would be identified during reconnaissance, confirming the observation that they are primarily constructed adjacent to the valley itself and along the pathways that traverse the immediate vicinity. One concentration of geoglyphs in particular formed the initial set of targets used to structure the ensuing UAV-pedestrian survey, with an additional control area selected that had yielded only 3 geoglyphs in our analysis of the World View 2 imagery. 3.2. UAV image acquisition and Pedestrian survey Areas having a high concentration of geoglyphs based on the initial assessment of the World View 2 imagery (as well as a control area) 1

0.0.

Moran's I index: 0.2470251, Expected index: −0.008065, z-score: 5.0442, p-value:

122


Fig. 3. Map of initially identified geoglyphs and other anthropogenic features targeted for reconnaissance and drone mapping.

were visited during a two-week period in the summer of 2015 (Fig. 4). A DJI Inspire 1 Quadcopter UAV was used to collect high-resolution photographs in multiple passes at various elevations, generating hundreds of images that were used to create detailed georeferenced orthophotos of each study area. 3.2.1. General observations after using drones for general mapping While drones provide a certain level of ease when it comes to data acquisition, two issues proved to be quite cumbersome in the day-today operations of the survey. The main issue for their use in survey was limited by the topography of the terrain to be studied. While most of the survey was conducted at the level of the pampa, initial attempts to fly the drone up to the desired location from the valley floor clarified the limits of the range of the radio transmitter that comes standard to this model of UAV. It was our understanding that the wireless capabilities of the drone would be sufficient for us to capture data remotely without having to climb the steep terrain ourselves. The topography of the terrain and proximity of targets to access points to the pampa proved to be more complicated than expected, with gusts of wind and rock outcrops endangering the UAV and the limited range of the remote connection. Therefore, the drone and all its components were taken onto the pampa by foot and features were recorded at a closer distance to allow for a more direct line of sight between the drone and the operator. This involved carrying a 12 kg case with components up a steep slope, at times over distances of more than a kilometer up the valley where considerably stronger winds were encountered. As this model of UAV had the option to use its GPS location to stabilize its position without operator control, the unit would automatically reposition itself again on a desired location regardless of windspeed, within limits. Although the specific model of UAV used was entirely satisfactory for this mapping project, there were a few minor limitations that

presented themselves during use in the field. Although admittedly minor and entirely specific to the DJI Inspire 1, the amount of time required to assemble and disassemble the drone prior to and after each flight was one concern. The second more major issue and fundamental limiting factor in terms of productivity is related to the longevity of each drone battery. With each battery pack providing a maximum of 15 min of flight time under optimal conditions with low wind speeds, this would allow for only 10–13 min of active data acquisition in order to save sufficient power to for the return journey to the takeoff location. In order to maximize daily productivity, a total of 6 batteries were purchased which would allow for approximately 2 h of active survey per day. In practice, depending on the size of the survey area, complexity of terrain, number of photographs taken as determined by the elevation of the flight for a specific desired resolution, one battery was often sufficient. For example, the Pampa de Siguas pass managed to cover 1 km2 with a single battery, while the Cujan Alta pass used 2 batteries and only covered 0.34 km2. In order to acquire imagery at a resolution that would suit the limits of our research question and maximize area covered, we found that the optimal elevation for our flights was between 50 and 75 m above the ground surface in transects that allowed for 80% overlap between pictures within and between each transect. As the X3 Zenmuse FC350 camera that comes standard with the DJI Inspire 1 has an effective resolution of 12.4Megapixels and a 94° field of view, we found that 80% overlap between images was more than sufficient for our purposes resulting in extremely high resolution final orthomosaic images of each survey area. 3.2.2. The georectification and Orthorectification of acquired imagery The individual photographs captured for each survey area were combined using the photogrammetric software package Agisoft Photoscan Pro (Version 1.2.0 build 2198) to create a single orthographic


123

Fig. 4. Map of areas surveyed and mapped using the drone.

image for each of the UAV survey areas (Fig. 5). Exact algorithms used by Photoscan in the processing of drone-captured imagery are unfortunately proprietary, but use a variety of Structure from Motion (SFM) and other computer vision algorithms (Verhoeven, 2011). Because the particular UAV used in this project was GPS enabled, the spatial information

for the location of each image was recorded within the metadata of each file, allowing the software to automatically position the resulting photogrammetric model within geographic space as per the local projection. With all photographs aligned based on their capture location, the collection of images were then combined as an orthomosaic to create a single

Fig. 5. Example of Agisoft Photoscan Pro workflow, showing a) Sparse Cloud, b) Surface, c) Textured, and d) Orthophoto.

124


Fig. 6. Diagnostic ceramic sherds photographed during the combined UAV and pedestrian survey include these Late Intermediate Period local style examples (left) from Quilcapampa Norte and these Tumilaca-style fragments (right) from Cujan Alta that date to the very end of the Middle Horizon.

high resolution image of the entire survey area. Comparison of the resolution of the resultant composite georectified orthophoto to ground truthed features of known size, it was calculated that each pixel of the composite image was equivalent to approximately 2 cm, allowing for recognition of comparatively small artifacts and features such as individual construction elements within walls or even single human long bones from looted tombs that were otherwise invisible in the satellite imagery. As the final high resolution composite file was georectified and orthorectified according to the appropriate Universal Transverse Mercator (UTM) projection within Agisoft Photoscan Pro prior to export to GIS, this resulted in a much more accurate final product and greatly reduced the post-processing time associated with a manual transformation of the image.

This material lends support to the Middle Horizon and Late Intermediate Period dating of geoglyph construction on the Sihuas pampa (Fig. 6). Combined pedestrian- and UAV-survey provided a very high resolution picture of geoglyphs and associated anthropogenic features. Many additional features that were recorded in the field had not been originally detected by our analysis using the World Viewe 2 satellite imagery. In the section that follows, we briefly discuss two areas in our survey area to illustrate the increased data capture of drone survey. These are the site of Cujan Alta (or P13) and a second area away from the valley that served as a control of the data capture quality (P10). Comparison between the two area revealed that dozens of smaller features were either missed or misinterpreted during early investigations of the satellite imagery, obscuring a complex ritual landscape that was only revealed during our two-week field season.

3.2.3. Tying aerial and pedestrian survey together The associated pedestrian survey allowed us to ground truth those features that we identified as petroglyphs in our initial analysis of the satellite imagery. The selected flight paths for the UAV survey contained a total of 30 geoglyphs identified based on the World View 2 imagery analysis, with additional architectural and unknown features also targeted; the latter were confirmed as a rock line and rock cluster respectively. Other unknown features were identified as contemporary anthropogenic features or have yet to be identified due to their inaccessibility. Ground-truthing also allowed us to look for diagnostic material associated with the geoglyphs that could buttress the Middle Horizon and Late Intermediate Period dates assigned to these features by other researchers based on rock art style alone. A non-destructive or non-extractive site survey methodology was followed and the little material that was found was only photographed in the field (Parcak, 2007).

4. Preliminary results of the 2015 Quilcapampa Geoglyph Survey Field Season The Quilcapamapa Geoglyph Survey prioritized the documentation of circular geoglyphs for the 2015 field season. Three major classifications were created during the initial assessment of archaeological potential on the pampa using World View 2 imagery to distinguish between their varied form based on the number of concentric rings (Fig. 7). Single ringed were the most abundant (n = 42), but were by far the smallest in terms of overall size (Fig. 8). Double ringed geoglyphs numbered less than half (n = 22) of single ringed geoglyphs, but were nearly twice the mean size. Lastly, multi-ringed geoglyphs, containing three or more concentric rings, were the least abundant (n = 14), but were by far the largest in terms of mean size. Within this last category it should be noted that there are a few very large examples.

Fig. 7. Three major geoglyph classification of a) Single, b) Double, and c) Multi-ring.


125

Fig. 8. General overview of Sihuas pampa geoglyphs, a) site size by typological classification, and b) site size frequency.

Cujan Alta (P13) was identified as a high-concentration geoglyph area based on the original satellite imagery (Fig. 9). In our initial analysis of World View 2 imagery, we identified 4 geoglyphs in the area of Cujan Alta. The area also contained one of the largest geoglyphs surveyed in the area, a multi-ringed geoglyph measuring c. 867 m2. Ground reconnaissance of Cujan Alta revealed a dense anthropogenic landscape of fortification walls, structures, tombs, and negative geoglyphs that were associated with the geoglyphs visible in the World View 2 imagery. Like the other high-concentration areas on the Sihuas pampa, Cujan Alta is located around a path climbing out of the valley. This path was later cut by the series of fortification walls, and there are other indications of activity, such as tombs, being placed over existing features. The earliest ceramic fragments found in the area are Tumilaca-style, dating to the very end of the Middle Horizon (Owen, personal communication 2015) (see Fig. 6). Like elsewhere in Peru (Reindel et al., 2006), the Sihuas geoglyphs were subjected to a complex array of treatments and associations as people continued to interact with the rock art long after they were made. Only 3 geoglyphs were observed in the satellite imagery within a second UAV survey area (P10) during our initial assessment of feature potential away from the valley itself. The area was chosen as a control to confirm that low-concentration area as identified via satellite imagery contained few anthropogenic features when assessed using UAVpedestrian survey. The results from P10 demonstrated that that this was indeed an area of low anthropogenic investment (Fig. 10), thus supporting our assertion that geoglyphs clustered tightly along the

valley rim. A fourth, previously unobserved geoglyph was nonetheless found following the drone mapping in the area. This newly identified, single-ringed geoglyph of 9 m2, is also the smallest one found in the valley so far. This instance highlights the differences in data quality available between high resolution satellite imagery and drone-assisted survey. Those few geoglyphs located in low-concentration areas were typically associated with inter-valley trails, as was the case with the geoglyphs found in this area which were less than a meter away from the road. These foot trails were the principle means of transportation across much of Peru prior to the construction of a national railroad system in the late nineteenth century (Klarén, 2000: 177). The combined UAV and pedestrian survey, not surprisingly, identified small-sized geoglyphs and other anthropogenic features that were not visible using the World View 2 imagery. Additionally, slope was another limiting factor that led us to miss from space some of the geoglyphs later found in different survey areas. This included a llama geoglyph (G081) from Timaran Alto (P01) (Fig. 11). This geoglyph was unique in that it was a very large built up or positive geoglyph, while all others were much smaller or composed in “campo barido” style. The pampa, especially along the valley's edge, is an undulating plain and some geoglyphs were placed on slopes that left them foreshortened when viewed from space. The three-dimensional photogrammetric model created by the UAV images—aided by observations during pedestrian survey—allowed us to manipulate the model to definitively identify and map these foreshortened features. Finally, there were features like parallel lines that were visible in the World View 2

126


Fig. 9. Detail map of Cujan Alta (P13) showing differences between a) WORLD VIEW 2 and b) drone image quality and c) mapping of major anthropogenic features.

images, but whose natural or anthropogenic origins were unclear. Many of these lines could be clearly identifiable as geoglyphs when we visited the area. 5. Conclusion Geoglyphs are found throughout the coastal plain of the central Andes. While attention is most often paid to the large zoomorphic figures among the Nazca corpus, most geoglyphs are smaller, simpler, and hence less well-documented. The Sihuas geoglyphs are dispersed across the pampa on both sides of the river valley. Some examples like the Gross-MUNSA have been previously reported, but most have been largely ignored in archaeological descriptions of the valley's history (Linares Málaga, 1990: 311–354; Linares Delgado, 2009). The Quilcapampa Geoglyph Survey was designed to gain a better understanding of the region's geoglyphs and their association with geologic and other anthropogenic features. Our approach uses a combination of conventional satellite imagery and new aerial mapping techniques to reduce the total time and effort associated with documenting a large number of features across such a vast expanse. Our initial season of fieldwork clearly demonstrates that the integration of satellite and UAV imagery can provide an unprecedented level of detail to aid in the analysis of large-scale human activity across broad landscapes. Time in the field was much more efficient because the satellite imagery was used as an initial step in identifying high-concentration areas of geoglyphs. Our analysis of the World View 2 imagery demonstrated that most petroglyphs were clustered around the trails leaving the valley, and then our further combined UAV/ pedestrian survey provides a far richer picture of these high-concentration areas so that additional features could be identified, mapped, and classified.

Our second season of field research continues to investigate the spatial relationships between different anthropogenic landscape features on the Sihuas pampa. With a better understanding of what geoglyphs and other features look like on the ground, we are currently improving our visual recognition of these objects in the World View 2 imagery. We are also mapping additional high-concentration areas using combined UAV and pedestrian survey, as well as adding more control areas where geoglyphs appear to be sparse or absent. Further classification of geoglyphs and other features in these high-concentration areas will provide us with a richer understanding of the ritual landscapes surround the Sihuas valley. Finally, we are also beginning to incorporate aerial photographs from Peru's Servicio Aerofotografico Nacional (SAN) into our database. The resolution of World View 2 images are often on par or superior to the SAN photographs, but the first images from the latter source are from the 1950s and may reveal petroglyphs that have subsequently been damaged and destroyed. A comprehensive full coverage survey of the Sihuas pampa—one supported with UAV technology—would undoubtedly uncover features that we have missed using the methodology presented here. Considerable effort, however, would be expended in areas where no rock art is likely to have occurred, and would require a well-funded, multi-year project to complete. The Sihuas geoglyphs, unfortunately, cannot wait this long. The Proyecto Especial Majes-Sihuas (Autoridad Autonomo de Majes, 2016) is moving quickly on an irrigation project to dam tributaries and pump water on to the pampa. This work builds off of earlier irrigation projects in the region, and is a growing trend throughout Peru. As the work continues, roads are built, lots are being bought, and water tanks are being constructed. A few of the pampa's geoglyphs have already been damaged, many are in jeopardy. By using a combination of satellite


127

Fig. 10. Detail map of control area (P10) showing differences between a) WORLD VIEW 2 and b) drone image quality and c) mapping of major anthropogenic features. Note the additional geoglyph identified by the drone imagery.

and unmanned aerial vehicle system (UAV, or drone) imagery within a Geographic Information System (GIS) to obtain high resolution images of some areas of the pampa, this project allows us to capture the most significant parts of a vast ritual—as well as economic and

political—landscape surrounding the Sihuas valley that will aid immensely in our reconstruction of past lifeways. More importantly, the images and maps generated from the Quilcapampa Geoglyph Survey are being shared with the Peruvian Ministry of Culture, assisting their

Fig. 11. Map of the Llama geoglyph (G081) from Timaran Alto (P01).

128


efforts to safeguard these features for future generations in the face of agricultural development. Acknowledgments The authors thank the following grant and funding sources for their generous contributions: Social Sciences and Humanities Research Council of Canada (Grant # 435150212), National Geographic Society (Grant # 9730-15), Royal Ontario Museum (Louise Hawley Stone Charitable Trust), and the University of Toronto Archaeology Centre. We would also like to thank Marcel Fortin and the Map and Data Library for purchasing the World View 2 imagery. Finally, we would like to acknowledge the Ministry of Culture of Peru for permission to carry out fieldwork in the Sihuas Valley (Resolución Directoral 218-2016/DPGA/ VMPCIC/MC). References Alexakis, D.D., Sarris, A., Astaras, T., Albanakis, K., 2009. Detection of neolithic settlements in Thessaly (Greece) through multispectral and hyperspectral satellite imagery. Sensors 9 (2), 1167–1187. http://dx.doi.org/10.3390/s90201167. Aveni, A.F., 1986. The Nazca lines: patterns in the desert. Archaeology 39 (4), 32–39. Aveni, A.F., 2000. Between the Lines: The Mystery of the Giant Ground Drawings of Ancient Nasca, Peru. University of Texas Press, Austin. Bahn, P., 2010. Prehistoric Rock Art: Polemics and Progress. Cambridge University Press, Cambridge. Ballarin, M., Balletti, C., Guerra, F., 2015. Action cameras and low-cost aerial vehicles in archaeology. In: Remondino, F., Shortis, M.R. (Eds.), Proc. SPIE 9528, Videometrics, Range Imaging, and Applications, p. 952813 http://dx.doi.org/10.1117/12.2184692. Barnes, I., 2003. Aerial remote-sensing techniques used in the management of archaeological monuments on the british army's salisbury plain training area, Wiltshire, UK. Archaeol. Prospect. 10 (2), 83–90. http://dx.doi.org/10.1002/arp.197. Beck, A., Philip, G., Abdulkarim, M., Donoghue, D., 2007. Evaluation of Corona and Ikonos high resolution satellite imagery for archaeological prospection in Western Syria. Antiquity 81, 161–175. Bescoby, D.J., 2006. Detecting Roman land boundaries in aerial photographs using radon transforms. J. Archaeol. Sci. 33 (5), 735–743. http://dx.doi.org/10.1016/j.jas.2005.10. 012. Casana, J., Cothren, J., 2008. Stereo analysis, DEM extraction and orthorectification of CORONA satellite imagery: archaeological applications from the near east. Antiquity 82, 732–749. Casana, J., Panahipour, M., 2014. Satellite-based monitoring of looting and damage to archaeological sites in syria. J. East. Mediterr. Archaeol. Herit. Stud. 2 (2), 128–151. http://dx.doi.org/10.1353/ema.2014.0011. Casana, J., Cothren, J., Kalayci, T., 2012. Swords into ploughshares: archaeological applications of CORONA satellite imagery in the near east. Internet Archaeol. 32. http://dx. doi.org/10.11141/ia.32.2. Ciminale, M., Gallo, D., Lasaponara, R., Masini, N., 2009. A multiscale approach for reconstructing archaeological landscapes: applications in Northern Apulia (Italy). Archaeol. Prospect. 16, 143–153. http://dx.doi.org/10.1002/arp. Conlee, C.A., 2003. Local elites and the reformation of Late intermediate period sociopolitical and economic organization in Nasca, Peru. Lat. Am. Antiq. 14 (1), 47–65. Contreras, D.A., Brodie, N., 2010. The utility of publicly-available satellite imagery for investigating looting of archaeological sites in Jordan. J. Field Archaeol. 35 (1), 101–114. http://dx.doi.org/10.1179/009346910X12707320296838. Cowley, D.C., Stichelbaut, B.B., 2012. Historic aerial photographic archives for European archaeology. Eur. J. Archaeol. 15 (2), 217–236. http://dx.doi.org/10.1179/ 1461957112Y.0000000010. Cox, C., 1992. Satellite imagery, aerial photography and wetland archaeology. World Archaeol. 24, 249–267. http://dx.doi.org/10.1080/00438243.1992.9980206 (April 2014). De Laet, V., Paulissen, E., Waelkens, M., 2007. Methods for the extraction of archaeological features from very high-resolution Ikonos-2 Remote sensing imagery, Hisar (Southwest Turkey). J. Archaeol. Sci. 34 (5), 830–841. http://dx.doi.org/10.1016/j.jas.2006. 09.013. De la Vera Cruz, P., 1996. El papel de la sub región norte de los valles occidentatales en la articulación entre los Andes cnetrales y los Andes centro sur. In: Albó, X., Arratia, M., Hidalgo, J., Núñez, L., Llagostera, A., Remy, M., Revesz, B. (Eds.), La integración surandina: cinco siglos después. Universidad Católica del Norte de Antofagasta, Arica, pp. 135–158. De Reu, J., Plets, G., Verhoeven, G., De Smedt, P., Bats, M., Cherretté, B., De Maeyer, W., et al., 2013. Towards a three-dimensional cost-effective registration of the archaeological heritage. J. Archaeol. Sci. 40 (2), 1108–1121. http://dx.doi.org/10.1016/j.jas.2012. 08.040. Fowler, M.J.F., Fowler, Y.M., 2005. Detection of archaeological crop marks on declassified CORONA KH-4B intelligence satellite photography of Southern England. Archaeol. Prospect. 12 (4), 257–264. http://dx.doi.org/10.1002/arp.266. Gheyle, W., Trommelmans, R., Bourgeois, J., Goossens, R., Bourgeois, I., De Wulf, A., Willems, T., 2006. Evaluating CORONA: a case study in the Altai Republic (South Siberia). Antiquity 78 (300), 391–403.

Goossens, R., De Wulf, A., Bourgeois, J., Gheyle, W., Willems, T., 2006. Satellite Imagery and Archaeology: the example of CORONA in the Altai Mountains. J. Archaeol. Sci. 33 (6), 745–755. http://dx.doi.org/10.1016/j.jas.2005.10.010. Hadjimitsis, D.G., Themistocleous, K., Agapiou, A., Clayton, C.R.I., 2009. Monitoring archaeological site landscapes in cyprus using multi-temporal atmospheric corrected image data. Int. J. Archit. Comput. 7 (1), 121–138. http://dx.doi.org/ 10.1260/147807709788549376. Hill, A., Rowan, Y.M., Kersel, M.M., 2014. Mapping with aerial photographs. Near East. Archaeol. 77 (3), 182–186. http://dx.doi.org/10.5615/neareastarch.77.3.0182. Klarén, P., 2000. Peru: Society and Nationhood in the Andes. Oxford University Press, New York. Kouchoukos, N., 2001. Satellite images and near eastern landscapes. Near East. Archaeol. 64 (1), 80–91. Lasaponara, R., Masini, N., 2006. On the potential of QuickBird data for archaeological prospection. Int. J. Remote Sens. 27 (16), 3607–3614. http://dx.doi.org/10.1080/ 01431160500333983. Linares Delgado, L., 2009. Reconocimiento arqueológico del valle de Sihuas. Bol. Lima 155, 51–70. Linares Málaga, E., 1978. Prehistory and Petroglyphs in Southern Peru. In: Browman, D.L. (Ed.), Advances in Andean Archaeology. Mouton Publishers, New York, pp. 371–391. Linares Málaga, E., 1990. Prehistoría de Arequipa. CONCYTEC-UNAS, Arequipa. Linares Málaga, E., 1999. Arte Rupestre en Sudamérica: Prehistoria. Fondo Editorial Universidad Mayor de San Marcos, Lima. Linares Málaga, E., 2004. Arte Rupestre en Arequipa y el Sur del Perú. Universidad Católica Santa María, Arequipa. Linares Málaga, E., 2013. Evaluación de recursos histórico-arqueoógicos del Proyecto Majes y área de influencia. Sector Siguas-Huacán. Bol. APAR 5 (17-18), 761–777. Mancini, F., Dubbini, M., Gattelli, M., Stecchi, F., Fabbri, S., Gabbianelli, G., 2013. Using Unmanned Aerial Vehicles (UAV) for high-resolution reconstruction of topography: the structure from motion approach on coastal environments. Remote Sens. 5 (12), 6880–6898. http://dx.doi.org/10.3390/rs5126880. Mark, R., Billo, E., 2016. Low altitude unmanned aerial photography to assist in rock art studies. SAA Archaeol. Rec. 16 (2), 14–16. Mascort-Albea, E.J., Jaramillo, J.R., Romero-Hernández, R., 2014. Analizando El Patrimonio Cultural a Través de Vuelos No Tripulados: Nuevas Estrategias de Investigación de Espacios Rurales Y Urbanos. Virtual Archaeol. Rev. 5 (11), 55–64. http://dx.doi.org/ 10.4995/var.2014.4181. Nieves, A.C., 2007. Between the river and the pampa: a contextual approach to the rock art of the nasca valley (grande river system, department of ica, peru) (PhD Thesis) University of Texas, Austin. Núñez Jiménez, A., 1986. Petroglifos del Perú. Panorama mundial del arte rupestre. 2nd ed. PNUD-USECO Proyecto Regional de Patrimonio Cultural y Desarrollo, Havana. Palmer, R., Fowler, M.J.F., 2013. In: Hanson, W.S., Oltean, I.A. (Eds.), Archaeology from Historical Aerial and Satellite Archives. Springer, New York, NY http://dx.doi.org/10. 1007/978-1-4614-4505-0. Pappu, S., Akhilesh, K., Ravindranath, S., Raj, U., 2010. Applications of satellite remote sensing for research and heritage management in indian prehistory. J. Archaeol. Sci. 37 (9), 2316–2331. http://dx.doi.org/10.1016/j.jas.2010.04.005. Parcak, S., 2007. Satellite remote sensing methods for monitoring archaeological tells in the Middle East. J. Field Archaeol. 32 (1), 65–81. http://dx.doi.org/10.1179/ 009346907791071773. Parcak, S., 2009. Satellite Remote Sensing for Archaeology. Routledge, London and New York. Philip, G., Donoghue, D.N.M.N.M., Beck, A.R., Galiatsatos, N., 2002. CORONA satellite photography: an archaeological application from the Middle East. Antiquity 76 (291), 109–118. Prins, A.B., Adams, M.J., Homsher, R.S., Ashley, M., 2014. Digital archaeological fieldwork and the Jezreel Valley Regional Project, Israel. Near East. Archaeol. 77 (3), 192–197. http://dx.doi.org/10.5615/neareastarch.77.3.0192. Reindel, M., Isla, J., Lambers, K., 2006. Altares en el desierto: las estructuras de piedra sobre los geoglifos Nasca en Palpa. Arqueol. Soc. 17, 179–222. Reinhard, J., 1985. The Nazca Lines: A New Perspective on Their Origins and Meaning. Los Pinos, Lima. Salvi, M.C., Salvini, R., Cartocci, A., Kozciak, S., Gallotti, R., Piperno, M., 2011. Multitemporal analysis for preservation of obsidian sources from Melka Kunture (Ethiopia): integration of fieldwork activities, digital aerial photogrammetry and multispectral stereoIKONOS II analysis. J. Archaeol. Sci. 38 (9), 2017–2023. http://dx.doi.org/10.1016/j. jas.2010.05.018. Sever, T.L., Irwin, D.E., 2003. Landscape archeology: remote sensing investigation of the Ancient Maya in the Peten rainforest of Northern Guatemala. Anc. Mesoam. 14 (1), 113–122. http://dx.doi.org/10.1017/S0956536103141041. Silverman, H., Browne, D., 1991. New evidence for the date of the Nazca lines. Antiquity 65 (247), 208–220. Smith, N.G., Passone, L., Al-Said, S., Al-Farhan, M., Levy, T.E., 2014. Drones in archaeology. Near East. Archaeol. 77 (3), 176–181. http://dx.doi.org/10.5615/neareastarch.77.3.0176. Swaminathan, N., 2013. Drones enter the archaeologist's toolkit. Archaeology 66 (6), 22. Szykulski, J., 2010. Prehistoria del Perú sur. Tambo Bolet. Arqueol. 2, 13–396. Trier, Ø.D., Larsen, S.Ø., Solberg, R., 2009. Automatic detection of circular structures in high-resolution satellite images of agricultural land. Archaeol. Prospect. 15, 1–15. http://dx.doi.org/10.1002/arp. Tung, T.A., Del Castillo, M., 2005. Una visión de la slaud comunitaria en el valle de Majes durante la época Wari. In: Olaya, C.C., Romero, M. (Eds.), Muerte y evidencias funerarias en los Andes Centrales: Avances y perspectivas. Universidad Nacional de Federico Villareal, Lima, pp. 149–172. Tung, T.A., Owen, B., 2006. Violence and rural lifeways at two peripheral Wari sites in the Majes valley of southern Peru. In: Isbell, W.H., Silverman, H. (Eds.), Andean Archaeology III. North and South. Springer, New York, pp. 435–467.

P. Bikoulis et al. / Journal of Archaeological Science: Reports 10 (2016) 119–129 Ur, J., 2003. CORONA satellite photography and ancient road networks: a northern mesopotamian case study. Antiquity 102–115. van Hoek, M., 2003. The Rock Art of Toro Muerto, Peru. Rock Art Res. 20 (2), 23–34. van Hoek, M., 2010. 'Tropy' Heads in the Rock Art of the Majes Valley. Perú: Exploring their Possible Origin. Rupestreweb (http://www.rupestreweb.info/trophy.html, accessed 8.02.16). van Hoek, M., 2013. The Carancha and the Apu. Rock Art in the Death Valley of the Andes. Rupestreweb (http://www.rupestreweb.info/carancha.html, accessed 8.02.16). Vaughn, K.J., 2005. Crafts and the materialization of chiefly power in Nasca. Archeol. Pap. Am. Anthropol. Assoc. 14 (1), 113–130. http://dx.doi.org/10.1525/ap3a.2005.14.113.

129

Verhoeven, G., 2011. Taking computer vision aloft - archaeological three-dimensional reconstructions from aerial photographs with photoscan. Archaeol. Prospect. 18 (1), 67–73. http://dx.doi.org/10.1002/arp.399. Verhoeven, G., Doneus, M., Briese, C., Vermeulen, F., 2012. Mapping by matching: a computer vision-based approach to fast and accurate georeferencing of archaeological aerial photographs. J. Archaeol. Sci. 39 (7), 2060–2070. http://dx.doi.org/10.1016/j.jas. 2012.02.022. Whitley, D.S., 2011. Introduction to Rock Art Research. 2nd ed. Routledge, London and New York.