Forensic Science International
Volume 183, Issue 1 , Pages e11-e16, 10 January 2009

Suspect burial excavation procedure: A cautionary tale

School of Geography, Archaeology & Palaeoecology, Queen's University Belfast, BT7 1NN Northern Ireland, UK

Received 21 February 2008; received in revised form 26 September 2008; accepted 20 October 2008. published online 10 December 2008.

Article Outline

Abstract 

Geographic location, time of reporting and need for rapid evaluation contributed to a lack of intelligence concerning a suspect burial site in scrub woodland (∼15km from the last known location of a missing person) in Northern Ireland. Police received reports of a subsiding ‘grave’, which was evaluated positively using GPR and victim recovery dogs (VRD). After 24h work, archaeological excavation showed a vertical-sided, stepped excavation on undisturbed clay with no inhumation. Subsequent research showed the feature to be an engineering trial pit. The GPR response was a water table and rocks, VRD were possibly reacting to disturbed ground. The work serves as a demonstration of good archaeological practice in suspect burial excavation, following a lack of landscape evaluation and poor overall intelligence.

Keywords: Forensic science, Clandestine burials, Victim recovery dogs, Ground-penetrating radar, Forensic excavation

 

Back to Article Outline

1. Introduction—why this is more than another clandestine grave case study 

Common practice [1], [2], [3], [4] in the use of forensic archaeology for the location and recovery of buried human remains, demonstrates how mapping (using topography, aerial imagery, remote sensing) is used to eliminate areas and focus walkover, geophysical-, dog- and probe-based searches. Together, these provide target(s) for detailed mapping, clearance and excavation for the recovery of human remains. In this case study, an area of depressed or subsident ground had features consistent with a clandestine burial (see Fig. 3.12—the ‘primary burial depression’ in ref. [1]). These included: the covert nature of the location, proximal to the last known whereabouts of a missing female; the size and shape of the collapse; ground-penetrating radar (GPR) definition of a dug trench with anomalies within and indications from a victim recovery dog (VRD, or cadaver dog). Such features are all considered by forensic archaeologists as good indications of a clandestine burial [1]. However, while some very realistic warnings of false-positive indications of burials are provided by experienced forensic archaeologists [2], including buried animals, the common engineering practice of trenching to examine ground conditions [4] creates features entirely consistent with a clandestine burial. This can include vegetational restoration of excavated ground, covering up of tracks made by mechanical diggers and limited disturbance to the surrounding environment. Hunter and Cox [2] shows how important walkover searches, dog reactions and geophysics are in finding and defining the limits of a clandestine grave. Hunter and Cox [2] demonstrate how in walkovers, scattered remains, animal scavenging and the ‘primary burial depressions’ of [1] are key to locating possible burials. Ref. [2] also include a specific chapter [5] on the use of geophysics (especially GPR) for finding anomalies but also for assisting in defining what those anomalies may be. In this case study, a burial-shaped depression, dog reactions and GPR were all consistent with the covert location of the feature, close to the last-known sighting of a missing person in Northern Ireland.

Back to Article Outline

2. Background 

2.1. Geographic location—description 

The site was located in an area of scrub woodland next to a small lake and to the rear of a retail park on the outskirts of an affluent conurbation in Northern Ireland. The adjacent land was under intensive development for additional retail amenities and housing. It was also some 15km from the last known location of a well-publicised “no body” murder victim case, reported in February 2005. Derelict and lowland areas such as this had received particular attention from the Police Service of Northern Ireland (PSNI) during their investigations. These were searched using trained personnel and victim recovery dogs (VRD), but with no indications of a clandestine burial having been discovered. The location is at the interface between countryside and town, and is frequented by dog-walkers, people engaged in covert alcohol consumption, and unlicensed fishing on the small lake. Access from the retail park to the site would normally be blocked by 2.5m-high steel fencing, but some of the metal bars had been removed to allow access by persons unknown. The area of interest was 75m from the gap in the metal fencing, along well-worn muddy and humus-rich woodland paths.

2.2. Time of reporting 

Local dog-walkers had noted an area of ground disturbance within this location in early 2006, approximately 2.5m east of a rough but frequently used path within the scrub woodland; immediately to the west of the path was a peat-filled, water-logged ditch, beyond which was the small lake. In May 2006 a member of the public had noted an area of subsidence within the disturbed area, measuring approximately 1.6m long×60cm wide and some 5–20cm in depth, and on a Friday morning notified the PSNI.

2.3. Preliminary investigations 

A PSNI Scene of Crime Officer was concerned that the area of subsidence, which was clearly identifiable as an infilled excavation (Fig. 1a), bore all the hallmarks of a clandestine burial [1], [2]. Bare soil with some grass and ivy growth surrounded the excavation to a radius of 1–2.5m. Beyond this, alder, holly, brambles, nettles, and bluebells increased in abundance, with more mature woodland (alder, sycamore, maple with honeysuckle, and blackthorn) present at a radius of 2–3m. As a consequence the officer requested the deployment of ground-penetrating radar (GPR), victim recovery dogs, and geographical profiling to ascertain the nature of the anomaly.

  • View full-size image.
  • Fig. 1. 

    (a) Area of subsidence prior to excavation, looking east. Measuring staffs are 1m long and were cleaned prior to use. The area has not been altered at this time and represents what members of the public and the first police officers at the scene observed. (b) Post-excavation view of cut feature with trench extension platform to the right of the picture, looking south-west.

GPR data were collected using two systems. Initial investigations were undertaken using a modified GSSI 2000 with shielded 225MHz antennae. These results were compared to those generated using a Mala RAMAC system with unshielded 200 and 400MHz antennae. The GSSI results showed a clear excavation, with a hyperbolic anomaly at 1.1–1.2m in depth (Fig. 2). GPR data from known burials commonly show a hyperbolic reflection or series of reflections from the human cadaver and skull [6], [7]. The Mala system also showed the limits of the excavation, with only minor changes in reflections at 1.1–1.2m in depth. GPR confirmed the two features known from other studies, that of hyperbolic reflections from a cadaver [6], [7], other objects within the collapsed ground, interference from out-of-plane objects such as tree roots and a collapse in surrounding soil [7]. Victim recovery dogs were deployed to the scene, where they gave an indication at the western end of the suspect excavation. A White's 750 (dual-loop deep, large target) metal detector showed consistent indications at the same location as the dogs. Given that this was a definite cut feature, and based on the results obtained during the deployment of the GPR and the victim recovery dogs, a decision was reached by the PSNI that they should immediately commission a team of archaeologists to undertake a forensic excavation, with work commencing on the Saturday morning.

  • View full-size image.
  • Fig. 2. 

    Ground-penetrating radar data from the area. These raw (unprocessed, field) data are displayed in order to demonstrate what was observed. The upper two images cross the suspect burial, the lower image is a control. As can be seen, the cut is clearly imaged with hyperbolae within. Data collected on a GSSI SIR200 GPR (modified for military use) using 225MHz shielded antennae, on a step-size of 5cm.

Back to Article Outline

3. The forensic excavation 

The area of subsidence was visible as a sub-rectangular cut feature, approximately 0.6m in width (north-south) and 1.6m in length (east-west) (refer to Fig. 1a). A grid was laid out over the area of subsidence with its long axis aligned east-west (Fig. 3). The trench was 3m long and 1.5m wide, and was sub-divided into eighteen 0.5m×0.5m grid squares, which were labelled from A to R (Fig. 4a). The excavation methodology was intended to facilitate the recovery of forensic evidence through the use of archaeological techniques. It was therefore necessary to implement a strategy of total recovery of all excavated material and all material removed from the cut feature was placed in large plastic sample bags assigned a unique bag number, with a total of 353 bags of soil collected during the course of the investigation.

The vegetation was removed and retained for each grid square across the trench; the loose soil on the surface of the area of subsidence was then removed and bagged. This work helped to define the edges of the cut feature, and demonstrated that it was a rectangular cut, some 2.0m long (east-west) and 0.7m wide (Fig. 4b). The initial phase of the excavation focused on the western half of the trench (grid squares J to R; Fig. 4a), as this was where the potential anomaly had been highlighted by the GPR survey. This strategy was implemented with a view to excavating the remainder of the trench if anything of significance was found in the western half. The fill in the western part of the trench was excavated in levels down to a depth of approximately 0.7m. At this point the trench became too deep to work in safely, and excavation therefore continued at the eastern end (grid squares A to I; Fig. 4a). The eastern part of the trench was brought down to the same level as the rest of the trench (approximately 0.7m).

The fill within the cut feature was a heterogeneous deposit that comprised mixed topsoil and re-deposited subsoil. The fill also contained some large stones (average size 200mm×100mm×50mm), but was remarkably sterile except for the discovery of two metal drinks cans recovered from Square N at the western end of the trench at a depth of approximately 30cm. Green leaves and fresh-looking shoots were recovered, however, from depths of up to 0.8m, thereby suggesting that the cut had been excavated relatively recently. Tree roots were also encountered, many of which had been cut, presumably during the original excavation of the feature. One large root, however, remained in situ between grid squares JKL and MNO (Fig. 4a). There were several possible tool marks present on the surface of the root, indicating that a failed attempt had been made to remove it from within the cut.

The water table was encountered at a depth of approximately 1.0m. At this stage a decision was made to alter the excavation methodology, and the trench was extended by approximately 1.0m×3.0m to the north in order to provide a level platform to work from (Fig. 5). This trench extension was excavated by PSNI officers using spades (Fig. 1b). The deposits in the extended area were found to be sterile and comprised compact brown loam topsoil, overlying a compact gravely clay. Both the topsoil and the clay were homogenous and extremely compact in comparison with the fill of the cut feature. Excavation within the trench continued, with a large stone (approximately 0.2m×0.2m) encountered at a depth of between 1.20 and 1.36m, and work terminated when the surface of a natural lacustrine grey clay was encountered at a depth of between 1.5 and 1.7m.

The excavation had revealed that the edge of the cut feature was well defined and there was no evidence for a re-cut. The base of the cut was broadly U-shaped in profile (Fig. 6a); the northern and southern sides were vertical, terminating in a flattish base (Fig. 6b). The eastern side was steeply sloped, with a sharp break in slope from ground level (Fig. 6a), while the western side was vertical to a depth of approximately 0.7m, and then sloped down steeply to a maximum depth of 1.5m (Fig. 6a). There was no inhumation discovered within the cut feature and its fill was uniformly sterile and contained no evidence (for example, maggots or staining of the soil by decomposing bodily fluids) to suggest that a body had ever been present within it.

Back to Article Outline

4. Conclusions 

This excavation revealed a genuine cut feature, but one that presented no evidence of having been excavated for the purpose of a burial. Here we go through the features of the site that led to the police commissioning a major excavation, and consider each feature, such that others can follow our reasoning, and see what problems arose.

4.1. Geographic location 

This fitted the profile of a covert, peri-urban location for a clandestine burial [3]. Vehicles could legitimately enter the retail area at any time of day or night, be driven to the rear of the property and be parked immediately adjacent to the gap in the metal fencing. The gap in the fence opened up to the base, making it wide enough to walk through or carry/drag objects through. The distance of 75m to the suspect burial site is within the range one or two persons could carry a corpse. The location of an illegal drinking den suggested this location was known as a place unlikely to be visited by people who may report activity to the police. The limited number of access points to the location made the possibility of people walking through the woods unlikely. The ground comprises soft loamy soils, clayey gravels and peat, all of which could be easily dug. The adjacent retail park is located on a major ring-road that circles a large, affluent town in Northern Ireland, where police road-blocks are infrequent. The road is on a direct route from the last-known location of a supposed no-body murder victim, who went missing 18 months prior to the site being discovered. In conclusion, the geographic location was an ideal burial site.

4.2. Time of reporting and area of subsidence 

The witness report of a subsident area came 18 months after the last-known location of a young woman 15km away. From witness reports the young woman was presumed to have been the victim of a suspicious death, most likely murder. Media attention on this case was high, with prominent posters along roadsides in the area. Television shows and local radio had been giving the case coverage in the weeks prior to the witness report. It could thus be claimed that local people were unintentionally looking for the remains or grave of the missing person, forming a kind of autosuggestion that a subsident path could be a grave. To support this, however, 18 months would be a reasonable period [1] of time for the cadaver to collapse and for soil to compact, forming the area noted. In conclusion, the timing of the witness report was entirely consistent with media attention and the time taken for subsidence to occur. The disturbed area measured approximately 1.6m long×60cm wide and some 5–20cm in depth, making it of very similar dimensions to those seen both in graveyard burials as well as clandestine locations [1], [2].

4.3. Ground-penetrating radar 

The two types of GPR system used to survey the site showed consistent results in the form of the cut and the hyperbolic anomaly. Ground-truthing showed GPR had successfully imaged the form of the cut. The anomaly detected by the GPR survey may have been produced by the effect of the water table on a large stone located at a depth of between 1.2 and 1.36m within the cut feature. The remains of the two metal drinks cans recovered within the fill of Square N were probably responsible for the readings obtained during the metal detector sweep. In conclusion, GPR data were entirely consistent with the feature being a grave with remains and other objects within.

4.4. Victim recovery (cadaver) dogs 

The reaction of the dogs to the suspect ground was one of the most definitive aspects to this investigation, yet also the most worrying for future work. Two case studies described by Hunter and Cox [2] provide some possible reasons for the dog reactions. In their Case Study 1 (‘Search for a young adult male in a wide rural environment’ [a man missing in the Falkland Islands]), Hunter and Cox (page 47) state “The dog responded in areas of peat cuttings at the edge of the settlement, but no remains were forthcoming and the response [of the dog] was eventually attributed to methane release from the peat bogs”. Their observation is backed-up by their Case Study 8 (‘Search for a victim of the so-called Moors Murderers’), the search for the buried remains of a child, one Keith Bennett. In this case, geophysical anomalies were vented for gas, the dog indicating at two locations. These were excavated, whereupon the dog's interest got less and less. Hunter and Cox state (page 57) “As in the Falklands, the dog appears to have been respecting gasses emitted from peat or vegetation within the peat.” In the present case study, we note that this location had been searched immediately following the last-known observation of the missing woman, with no dog reactions, yet when visited by the same dogs, 18 months later, but most critically, after the suspect cut had been made (note, in peaty soil) the dogs reacted to the disturbed ground. A similar problem to the false-positive reaction of the cadaver dog has been reported for carbon isotopes in water [8], wherein isotope anomalies consistent with buried remains were found (upon excavation) to be previously dug ground, again in a peat-rich soil.

Back to Article Outline

5. Conclusions 

Can an explanation be suggested for the presence of this cut feature within the scrub woodland? The dimensions, depth and morphology of the cut feature would suggest that it had been excavated using a mechanical excavator. In advance of a new development, it is standard practice for the civil engineering company employed to sub-contract a geotechnical investigation company to determine the level of the water table [4]. This is typically achieved by drilling boreholes using a drilling rig and it is a rigorous, but costly, methodology. A more cost-effective option is to simply employ the services of a mechanical excavator to dig a trial pit down to the level of the water table. An added benefit to this methodology is that it can also allow an engineer the opportunity to go down into the trial pit and view its strata, thereby enabling a visual assessment to determine the soil type and therefore calculate a safe bearing pressure for foundations at the required depth. A typical trial pit would be 1.0m×2.0m in dimension, and between 1.0 and 5.0m in depth, depending on the size of the machine being used, and the depth at which the water table is encountered.

The similarities in the dimensions between these types of trial pits and those of the cut feature excavated in the scrub woodland should be noted, as too should the east-west profile of the cut feature which bears the hallmarks of having been created by the east to west movement of the arm of a mechanical excavator positioned on the pathway to the west of cut feature (Fig. 1, Fig. 5). It can be envisaged that the excavator's bucket dug into the soil at the east end of the cut, dragging the bucket and its content towards the west end of the trench (and hence the gently sloping eastern profile), before the arm and its filled bucket were raised up, thereby creating a more steep-sided, almost vertical, western end to the trench. Finally, it should be noted that there is a new development taking place to the east of the wooded area, and it can be suggested that the cut feature was opened at the request of a civil engineering company employed on this building programme (coincidentally) at some stage after the disappearance of the missing young woman.

Back to Article Outline

References 

  1. Dupras TL, Schultz JJ, Wheeler SM, Williams LJ. Forensic Recovery of Human Remains. New York: Taylor & Francis Group; 2006;232 pp.
  2. Hunter J, Cox M. Forensic Archaeology: Advances in Theory and Practice. London: Routledge, Taylor & Francis Group; 2005;233 pp.
  3. Hunter J, Roberts C, Martin A. Studies in Crime: An Introduction to Forensic Archaeology. London: Routledge, Taylor & Francis Group; 1996;174 pp.
  4. Roskams S. Excavation. Cambridge: Cambridge University Press; 2001;310 pp.
  5. Cheetham P. Forensic geophysical survey (Chapter 3). In:  Hunter J,  Cox M editor. Forensic Archaeology: Advances in Theory and Practice. London: Routledge, Taylor & Francis Group; 2005;p. 62–95
  6. Owsley DW. Techniques for locating burials, with emphasis on the probe. J. Forensic Sci. 1995;40:735–740
  7. Nobes DC. The search for “Yvonne”: a case example of the delineation of a grave using near-surface geophysical methods. J. Forensic Sci. 2000;45:715–721
  8. Ruffell A, McKinley J. Geoforensics. Chichester: John Wiley & Son; 2008;332 pp.

PII: S0379-0738(08)00418-0

doi:10.1016/j.forsciint.2008.10.013

Forensic Science International
Volume 183, Issue 1 , Pages e11-e16, 10 January 2009

Article Tools

* Image Usage & Resolution