Responding to Holes at Fifa, a Case Study Using Drones

The first phase of the FTP project mapped the site and carried out oral interviews with those who interacted with the cemetery site of Fifa. This 2011 field season was successful in some respects as well as highlighting some of the shortcomings associated with the project methods.^{Footnote 34} One such deficiency was the discrepancy noted between the pre-season analysis of Google Earth^© satellite imagery of the site, used to quantify the number of looted graves, and the realization that not every hole identified from space is an actual ancient tomb.^{Footnote 35} Standing in the lunar landscape of Fifa, comparing satellite images (with limited resolution) with what was on the ground, led to the radical idea of using drones to gather data with much greater spatial resolution, which would allow for the more accurate identification of individual pits and the potential differentiation of unsuccessful looters’ pits from actual looted graves. It was in this way that the Landscapes of the Dead (LOD) research project was born. To this end, drones—both fixed and rotary wing—are being used to build a (w)hole picture of the mortuary landscape, including looting activity and holes.

As part of a comprehensive approach to the Early Bronze Age IA mortuary landscape at Fifa, and in cooperation with the Jordanian DoA under the umbrella of the FTP, the LOD project is using unpiloted aerial vehicles (UAVs), or “drones” as they are commonly known, to study the scale and pace of natural and cultural landscape modification at Fifa. The integrated approach of the LOD project combines low elevation aerial photography, and the resultant spatial data for digital mapping, pedestrian survey, and ground truthing, with oral interviews with the constituents who interact with the site and its objects. This innovative interdisciplinary and collaborative endeavor provides a thorough change-over-time examination of the effects of looting and landscape modification on the local community and the archaeological record, while, at the same time, assessing the impact of the Jordanian DoA’s anti-looting campaigns, national laws, and local community outreach programs.

Mindful of recent backlash against the use of drones, remote sensing, and satellite imagery—what archaeologist William Caraher refers to as a type of “technological solutionism—new-fangled technologies as a key to solving some of the ills affecting archaeological fieldwork”^{Footnote 36}—the LOD project sought to overcome the impersonal nature of “eyes from the sky,” which can have a dehumanizing effect wherein the landscape is considered to be a part of a scientific investigation devoid of local interaction.^{Footnote 37} The LOD project, of which drones are only one part, includes multifaceted modes of research and analyses, community engagement, and knowledge sharing. This unique program provides a methodology for better understanding the ancient and modern interactions with the site. Implicit and explicit in this approach is an acknowledgment of, and a focus on, the ethical implications of operating UAVs and the potential negative and positive impacts of this research on the local communities of the region, an often missing element from many research projects.^{Footnote 38}

The merit of using drones (ease, efficiency, and affordability) to document landscape features,^{Footnote 39} archaeological excavation areas,^{Footnote 40} and subsurface (thermal and near infrared) surveys^{Footnote 41} is well established, but the need for ground-truthing sites is echoed by most, if not all, practitioners who utilize satellite imagery, drones, or both for landscape interpretation and prospection. Pedestrian surveys are a vital component of archaeological landscape assessment;^{Footnote 42} without pedestrian ground surveys, features are undoubtedly missed or misinterpreted. In 2013, through comprehensive ground truthing and archival research, and in collaboration with a local archaeologist, an important discovery was made regarding the occurrence of looting, which was not identified using Google Earth^© or even with the unprocessed drone images. Many of the Early Bronze Age cist tombs share a single wall, a characteristic building technique confirmed by the earlier excavations of Mohammed Najjar (Figure 4).^{Footnote 43} Even where looting is most concentrated at Fifa, there are many undisturbed graves covered by the back-dirt piles of previously looted adjacent graves. However, the graves often share a wall, a phenomenon of which the looters are well aware. In an efficiently destructive approach, pristine graves can be rapidly accessed by digging sideways through the wall of a previously looted grave. It is much easier to dig sideways between two graves than to dig down from the surface through the overburden and the large limestone capstones covering the tombs.

Figure 4. Image of adjacent cist tombs sharing a single wall (courtesy of Morag M. Kersel).

The LOD project undertook aerial drone survey at Fifa from 2013 to 2016. Over the course of four seasons, as technology improved, a variety of fixed-wing and rotary-wing drones were used. The aerial surveys were designed to record the site at centimeter-level resolution in order to generate undistorted orthophotographs (see Figure 2) and digital elevation models (DEMs) of the site. After photogrammetric processing using Agisoft Photoscan Pro, the resulting data were analyzed via visual inspection and DEM comparison in order to identify new destruction at the site each year (Figure 5). Comparisons between 2013 and 2014 change over time, using the difference between annual DEMs, indicate discernable modifications to the landscape in the form of new holes as well as new positive change in elevation due to the creation of spoil heaps (Figure 5A), which are also visible in the orthophotographs as fresh (moist) dirt (Figures 5B and 5C), from new looting episodes, often alongside visible new pits (Figure 5D). Although, initially, it was unclear why there would occasionally be new back dirt adjacent to a previously looted cist grave, as mentioned above, pedestrian survey confirmed that sideways digging between looted and unlooted tombs created new back-dirt piles (positive change) without obvious new holes (negative change), a trend confirmed in later seasons (2015 and 2016).

Figure 5. Evidence for a single new hole: (A) difference in elevation between 2013 and 2014 (positive change is white, negative change is black); (B) orthoimage from 2013; (C) orthoimage from 2014; (D) hill shaded digital elevation model showing shape of new hole (courtesy of Austin (Chad) Hill).

One incentive for using drones to record change-over-time data is not only to examine the transforming landscape but also to provide comprehensive maps for use by the DoA, heritage organizations, and local communities, which may assist in the greater awareness of archaeological site destruction.^{Footnote 44} The high resolution diachronic data generated from the air, in conjunction with archival research, pedestrian surveys, site mapping, and oral interviews on the demand for, and the movement of, looted material are used to evaluate site protection strategies and to develop local community outreach programs, resulting in a comprehensive model that might be applied to other at-risk sites and regions.

Results of the Archaeological Response to Holes

There are a number of important results of the site monitoring with UAVs over four seasons of investigation at Fifa. First, despite the continued efforts of the DoA and local community programs of the Petra National Trust (PNT), looting persists. The individual reasons behind looting vary, but, in the final analyses of the trade in undocumented antiquities, the driving force is demand; institutional and individual demand for ceramic vessels from the Early Bronze Age results in the looting at Fifa.^{Footnote 45} Investigating this demand—another element of the FTP project—is a topic for another article, but there is compelling evidence that demand is driving the ruin of this (and other) landscapes.^{Footnote 46} As a component of the overall picture of site destruction, looting, and demand for archaeological artifacts, conducting oral interviews with various actors (dealers, looters, tourists, collectors, and intermediaries) provides an increased understanding of the constituent elements in the illegal trade in Early Bronze Age material from the Dead Sea Plain in Jordan, thus supplying complementary data to the UAV flyovers. Demand is for the ceramic vessels alone, not the other associated grave goods from the cist tombs.^{Footnote 47}

A second outcome of this research is an understanding of what remains at the site, which at first glance appears to be completely decimated. Though looting at the site continues to have a devastating effect, it would be imprudent to believe that the cemetery at Fifa is completely destroyed. In the area excavated by Mohammad Najjar in 2001, the density of tombs is estimated at approximately 25 graves per 100 square meters. The area covered by the identified boundaries of the site is approximately 70,658 square meters. Assuming that the density of burials across the entire site is similar to the results of this methodically excavated area, then the total number of burials at Fifa could be as high as 17,664. The 3,723 documented looting pits might then account for only 21 percent of the estimated total burials at the site. While this simple calculation likely overestimates the total number of graves at Fifa, even in the most disturbed part of the site, the number of successfully looted pits rarely approaches a density to match that of the scientifically excavated area of 2001. In these densely looted areas, there are between 5 and 10 looters’ pits per 100 square meters. Thus, even in heavily damaged areas, there may still be twice as many undisturbed graves as looted ones. There are many graves left to loot, and parts of the site remain intact.

The following is a detailed description of the quantitative and qualitative outcomes of a thorough methodology, which allows the LOD project to gain valuable insights into landscape change and the economic possibilities generated as a consequence of the holes at Fifa. Working collaboratively with the local authorities, the DoA, and NGOs allows for the exchange of information on these elements of looting in the Dead Sea region, which may aid in future anti-looting initiatives.

Results from Season 2013–14: New Holes, New Insights

Thirty-four new pits were dug between the 2013 and 2014 field seasons (Figure 6). While this is only a small proportion of the total holes at the site (less than 1 percent), it still represents a significant number of looted artifacts and potential prices realized. If each grave might contain an estimated 6–30 ceramic vessels, then the total number of saleable vessels from these 34 graves could be as low as 204 and as great as 1,020.^{Footnote 48} In today’s market, these pots sell for between $30 and $150, resulting in a low total estimate of $6,120 and a high estimate of $153,000 (see Table 1). Oral interviews with local looters (Looters 7, 8, 12) confirm that they receive between three and five Jordanian dinars ($4–7) per pot, which means they could be making roughly between 1.5 percent and 15 percent of the final price paid by the consumer.^{Footnote 49} This estimation corroborates the findings of Neil Brodie and Daniel Contreras who suggest looters make about 1 percent of the final sales price,^{Footnote 50} and the work of Jerome Rose and Delores Burke with looters in northern Jordan who earned about 15 percent on the sale of illegally excavated items.^{Footnote 51} In high and low estimates, looters could realize between $816 and $7,140 by illegally excavating 34 holes. The average yearly salary in Jordan is $5,160,^{Footnote 52} and while there is no suggestion that a single looter or even a group of looters is reaping all of the financial rewards, looting at Fifa has the potential to be a modestly lucrative endeavor.

Figure 6. Map with new looters’ pits 2013–14 (courtesy of Austin (Chad) Hill).

Table 1. Quantifying looted holes at Fifa

The 34 new holes were concentrated, not surprisingly, in the area of the main wadi (river bed) that runs through the site and down along the lower slopes in the southwestern edge—areas less visible from the road. Two types of looting are clearly evident at the site: systematic and opportunistic. Empty plastic water bottles, broken tools, sardine cans, and the detritus left by professional looters suggest individuals choose locations in the wadi or on the southern slope in order to dig in anonymity.^{Footnote 53} During this period, looters also methodically revisited old holes in order to dig in a sideways fashion as described previously. Archaeological and ethnographic data indicate there is also ad hoc looting by bored individuals with nothing to do, digging around in an already heavily looted landscape.^{Footnote 54} Evidence in the form of broken liquor bottles suggests that the lower slope to the southwest is a favored haunt for local youth, resulting in coincidental, rather than targeted, looting. Whatever the motivation, artifacts are purchased by intermediaries from the larger cities of Kerak and Amman and then traded through the extensive national and international networks (Dealers 17, 49).

Results from Season 2014–15: A Decrease in Looting Activity

An assessment of the maps between 2014 and 2015 indicate just three new holes during this period, which were analyzed using the same set of criteria that generated the following statistics (Figure 7). The estimated number of saleable pots from the three new graves is between 18 and 90. Today’s market prices of $30–$150 per pot provide a low total estimate of $540 and a high estimate of $13,500. Based on oral interviews with looters, the potential income generated from the illegal excavation of three holes could be between $72 and $630. Compared with the previous and subsequent years of analyses, there were fewer holes during this period, which results in questions surrounding the factors that are leading to less looting at Fifa. Between 2014 and 2015, a series of public outreach programs carried out by the PNT, in collaboration with the team from the FTP project, formed part of an all-encompassing approach to addressing issues related to archaeological site looting and the demand for artifacts (see the section on the response of NGOs to holes below).^{Footnote 55} Staff from the Jordanian DoA Safi Inspectorate and the Museum at the Lowest Place on Earth provided greater oversight and management of the site, with more on-site foot patrols and drive-by site monitoring. Moreover, a third possibility was uncovered through ethnographic interviews with buyers (Tourists 12, 24, 113) and sellers (Dealers 4, 8, 17, 29, 47, 49) of Early Bronze Age ceramic vessels for sale in the legal market in Israel. During the summer of 2014, there were fewer tourists in the region due to the Israel–Gaza conflict, which occasioned a downturn in artifact sales in general. Interviews with dealers also indicated that there was no demand for the Early Bronze Age pots, a consequence of changing consumer tastes (Dealers 4, 8, 17, 29, 47, 49). Russian icons were, and still are, the new “hot” item for religious pilgrims and other tourists (Dealers 4, 18, 23, 47). Are fewer holes at Fifa the result of effective outreach efforts, changing consumer tastes, or the downturn in regional tourism? Are our results skewed based on the time of year we surveyed the site (earlier in 2015 than previous or subsequent years) versus the potential peak season for looting, which is between harvests (see Kersel and Chesson Reference Kersel and Chesson2013b). Or is this season an anomaly?

Figure 7. Map with new looters’ pits 2014–15 (courtesy of Austin (Chad) Hill).

Results from Season 2015–16: Looting Activity Returns

In the analysis of aerial survey data between the 2015 and 2016 seasons, 24 new holes were recorded (Figure 8). Mapping the holes revealed an interesting pattern of looting. Along the northern extent, there are four new holes beyond the edge of the site, which were undoubtedly unsuccessful excavations. Looters are moving beyond the previous areas of concentration in the hope of finding new and untapped parts of the site. However, previous looting appears to have already effectively identified the margins of the cemetery. New holes are also apparent in formerly looted and scientifically excavated areas. A new cluster of holes is evident to the west of the 2001 scientific excavations of Mohammad Najjar (Figure 9). In the back-dirt pile of this looting, a complete vessel was recovered, suggesting perhaps novice looters or a certain degree of sloppiness in technique (Figure 10). Interactions with looters at the site have included groups of kids (aged between 8 and 16), seasoned regulars, and bored teens, who are mentioned above. The varying levels of experience and expertise in looting are evident in the tools left behind by the looters, the location of looting holes, the artifacts left behind, and the methods visible through ground truthing.

Figure 8. Map with new looters’ pits 2015–16 (courtesy of Austin (Chad) Hill).

Figure 9. Close up of orthophotographs demonstrating looting near the 2001 excavations: (A) 2015 and (B) 2016 (courtesy of Austin (Chad) Hill).

Figure 10. A complete discarded pot (courtesy of Austin (Chad) Hill).

Applying the same model of analysis based on patterns of grave goods identified by Rast and Schaub in the 1990 excavation at Fifa, the 24 new holes produced the following figures. An estimated low number of 144 pots to a high of 720 pots might have been recovered from the 24 holes, which could generate anywhere from a low total of $4,320 to a high of $108,000 in estimated sales. The potential income earned by the looters of these 24 holes is estimated to be between $576 and $5,040, which would be meaningful sums for the local residents of the Ghor es-Safi region. The 24 new holes recorded between 2015 and 2016 demonstrate an increase in looting in comparison to the 2014–15 season, where only three new holes were documented. The results from the 2014–15 season remain an irregularity.^{Footnote 56} Further investigation into the reasons for the low number is ongoing and could include the local economy, artifact demand, trade networks, and regional instability. Some general observations are possible based on the four seasons of drone flyovers (see Table 1). Analyses of the maps in conjunction with pedestrian surveys indicate that the looting of 61 holes at Fifa since 2013 has the potential to realize a total financial gain from sales in the range of $10,980 to $274,500. This does not reflect the prices paid to the looter (typically between $4–7 per pot), which could result in profits from looting between $1,464 and $12,810. Nor do these figures factor in how many hands the artifacts pass through in the associated financial transactions leading to the eventual purchasers; they only reveal the number of potential looted pots and the final sales price to the buyer. Quantifying a lifetime of looting at Fifa provides some stunning statistics from a single site (Table 2). Between 1989 and 2016, the looting of 3,723 graves potentially produced between 22,338 and 111,690 pots, which may have sold for between $30 and $150 per pot, resulting in prices realized between $670,140 and $16,753,500. Looters, paid anywhere from 3–5 Jordanian dinars ($4–7) for each pot, could possibly have earned 66,600 Jordanian dinars and 555,000 Jordanian dinars ($88,800 and $777,000).

Table 2. Quantifying a lifetime of looting at Fifa

Consumer tastes for items with a biblical association never completely go out of fashion, as attested by 27 years of looting in order to meet demand in the marketplace.^{Footnote 57} People buying Early Bronze Age pots from this region are typically tourists and religious pilgrims to the area who want to leave with a small memento of their trip.^{Footnote 58} The pots are prosaic, inexpensive keepsakes manufactured in the past that evoke a message about an associated biblical place. Bab adh-Dhra’ has been identified not merely as a city of the plain but also as the biblical Sodom, a city synonymous with sin. According to data collected, it is these biblical associations that make artifacts from the Dead Sea Plain sites desirable. Dealers interviewed during the summers of 2016 and 2017 indicated that they were already seeing a rise in demand for objects (including Early Bronze Age pots) with connections to the Holy Land (Dealers 6, 8, 23, 47, 49). The antiquities market is complicated, with illegal Jordanian artifacts legitimately available for the willing consumer in Israel, on eBay, and in other markets.^{Footnote 59} The desire for Early Bronze Age pots, as the drone survey and the oral interviews demonstrate, results in more holes. Given the continual demand for Holy Land material, the question of why there were only three holes between 2014 and 2015 remains. Perhaps recent NGO response to holes could be a factor in the fewer holes at Fifa?