Geophysical Investigation Using Ground Penetrating Radar (GPR) to Detect Unmarked Burial Sites Taher H 1

1, 2 Ameen ,

Rauf R

1 Hussein ,

Hanan H

1 Mahdi ,

Haydar J

1 Al-Shukri

University of Arkansas at Little Rock, USA. 2 University of Sulaimani, Iraq ([email protected], [email protected])

1. Abstract

4. Interpretation

Comprehensive Ground Penetrating Radar surveys were conducted at the Old Carlisle Cemetery, east of Little Rock, Arkansas, to investigate the locations of historic burial sites and to identify unmarked graves. The Old Carlisle Cemetery, Arkansas, since 1872 has been in use and a potential expansion will be planed if the geophysics data help to identify unused lands. GPR survey was conducted at the cemetery using GSSI SIR-3000 with 400 MHz and 900 MHz antennas and a tricycle cart. A total of 310.4m profiles of GPR data were acquired from three locations within the old and new parts of the cemetery. At the 1st location, GPR data collected along 6 parallel profiles. Data reveals, after the normal comprehensive processing; two unmarked graves at about a depth of 1m and one misplaced headstone or collapsed grave were detected. Other marked graves around 1m depth with headstones were also verified by their typical reflections hyperbola on the GPR profiles. At the 2nd location, the data collection was performed along 4 parallel profiles to locate potential areas that were not used for burial in the past. The GPR data showed that there were no graves in the area below at least two of the profiles. Three marked graves which were verified by their headstones might have metal caskets due to their strong reflection hyperbolas around a depth of about 1.2 m. Three other graves were either collapsed or decomposed due to their very weak reflections within a subsided surface area. Animal burrows and a rusted old key were found and verified by near surface digging. At the 3rd location, the data was collected along 3 parallel profiles. The GPR was able to detect one unmarked grave and two marked graves, each with two coffins, by showing strong reflection hyperbolas at about 0.75 m depth. A grave with a headstone to the north of the two graves did not show strong reflection hyperbola although the burial date (1987) is younger than the other two. This might reflect different type of burial practice or material than the other two graves. GPR did not detect any additional burial site at the studied locations.

Since the data acquired at three different locations within the cemetery, we expected to have different electrical properties for each area depending on the soil types and conditions. Hence the dielectric constant for each area was calculated separately.

2. Introduction The Old Carlisle Cemetery located in Carlisle about 38 miles to the east of Little Rock, Arkansas 72024. We have been approached by the Old Carlisle Cemetery Association Inc. which is a nonprofit organization that manages the cemetery, to help them in planning the extension of the old cemetery by conducting GPR surveys. Creek

Problem:  The north-east corner of the cemetery has been encroached by large trees and private hedge. It is believed that a grave may be outside the cemetery boundary and there may be unmarked graves in line with burial site no. 10 (in area 1).  Looking for empty lots to the west of area 2, and confirming if there are two adjoining grave plots that are close to burial site no. 17 (in area 2) near east edge of the cemetery.  Assuring that there is no burial sites under the recently planted headstones for future burials in area 3.  Search the 3 areas for unmarked or unknown graves and locate any misplaced stones.

N

0

44m

Fig. 1. Survey Layout map of the GPR profiles in Old Carlisle Cemetery showing the locations of area 1 (red), 2 (yellow), and 3 (green) profiles.

3. Data Acquisition

In area 1; Six parallel profiles trending N-S were acquired each 17.7 meters, except profiles no. 1 & no. 2 are20.8 m long.  In area2; Four parallel profiles trending N-S acquired each 17

Fig. 4. A photo viewing N showing survey area 3 with location of profiles 11 – 13. Area 1 Used Frequency

m long.

Interpreted data of profile no. 1: Identified burials with good hyperbolic reflections consistent with all the current headstones. Headstones; (2, 6 & 7) shifted by 0.5 - 1.0m. Some shallow reflections, expected to be the effect of the tree roots that are very close to the profile location. Interpreted data of profile no. 4: Unmarked grave; distance 5.0m, depth: 0.8m (18.5 ns). Identified burials; distance 6.0m, depth:0.85m (20ns). Burial #9 of six days old baby could not be verified; possibly very shallow within the grave shaft area. There is no evidence of burials after 7.0 m distance as the subsurface soil not disturbed. Interpreted data of profile no. 6: Identified burial at distance of 1.25 m and depth 0.85m (19ns), consistent with burial no. 10, its headstone shifted 1.5 m to the south. Two shallower hyperbolas close to burial site no.10 at depth of 0.5m (12ns), might represent the unmarked graves of the other two expected burials. Unmarked grave at distance of 8.0 m, and depth about 0.95m (21.5ns). No clear reflection for burial no.13 at distance 10 m, either has a deteriorated coffin or the headstone misplaced from the previous hyperbola considered as unmarked grave at distance 8.0 m.

Fig. 5. photo showing location of profile no.1; viewing South (Left), and Interpreted GPR data along the same profile with the location of burials relative to current headstones (Right).

Fig. 6. location of profile no.4 and no.6, with indicating unmarked graves (Left) and Interpreted GPR data along profile no.4 (Right).

Area 2

Area 3

GPR Profiles

Length (m)

GPR Profiles

Length (m)

GPR Profiles

Length (m)

1

20.8

7

17.0

11

17.9

2

20.8

8

17.0

12

17.5

3

17.7

9

17.0

13

17.5

4

17.7

10

17.0

5

17.7

6

17.7

profiles trending N-S each

14

20.8

18

17.0

17

18.5

900 MHz

15

20.8

16

17.7

acquired with 17.5 m long.

Total length

154

Total length

85

Total length

71.4

Total Length of all the profiles in the three areas

Fig. 3. A photo, showing survey area 2 with location of the profiles (7 – 10), viewing NW

Interpreted data of profile no. 13: Two marked graves, each with two coffins, showing strong reflection hyperbolas at distances of 2.5m and 12.5m and depths about 0.85m (11.5ns) and 0.65m (9.0ns) respectively. The horizontal soil layer is interrupted in the grave shaft areas, and it is undisturbed where there are no graves. A headstone at distance of 15 m; burial no.25 (1987), identified with a weak hyperbolic reflection at depth of 0.75m (10.5ns).

Fig.12. Interpreted GPR data along profile no.11

Fig.13. Interpreted GPR data along profile no.13

Fig.14. GPR data collected with 900MHz antenna, along profile no. 13, from N - S.

Fig. 7. Two unmarked graves (Left), and Interpreted GPR data along profile no.6 (right)

5. Summary

Fig. 8. GPR data of profile no.7; before migration (Left) and after migration (Right).

Fig. 9. Interpreted GPR data along profile no.8 (Left), and a Photo showing area 2 and the researchers (Dr. Hanan Mahdi & Taher Ameen) confirming GPR data on it; viewing NE (Right).

310.4 m

Table (1); GPR profiles acquired in the three areas of Old Carlisle Cemetery

Interpreted data of profile no. 11: Marked grave; distance:1.25 m, consistent with burial site no. 19, depth about 0.60m (8.0ns). Unmarked grave; distance: 2.5 m and depth about 0.50m (6.5ns) . A metal headstone of burial no. 20, distance (3.75) m, confirmed with a hyperbolic reflection at depth of 1.30m (18ns). No other burials along this profile, as a clear reflection of a horizontally soil layer is not disturbed.

Collected data with 900 MHz antenna: We used the 900 MHz antenna on the same profile no.13, hoping to see more details of the shallow reflections, but the results was not as remarkable as the 400 MHz antenna.

Verified by the depth of an old rusted key found along one of the profiles (profile no. 9) and the migration procedure on the best hyperbolas, a dielectric constant value of (6.7) was chosen for this area. Interpreted data of profile no. 8: A headstone at distance of 2.5m of burial no.14 (1942) confirmed by a very clear hyperbola at depth of about 1.15m (20ns). No clear hyperbolic reflections for three headstones;15 (1937) & 16 (1936), and 18(1936). A large headstone at distance of 8.0 – 9.6 m for burials 17a (1967) and 17b ( 1974) confirmed by two clear reflections at distances of 8.0m and 9.0m and depth of about 1.30m (22.5ns) and 1.10m (19ns) respectively.

A flat laying reflection was clearly identified along most of profile no. 11, type and depth of the interface causing the reflection was confirmed by trenching. A wet clay layer at 51 cm depth was found. The two-way travel time of GPR data is 7.0 ns resulted in a value of 4.25 dielectric constant for the dry soil layer (the upper layer). Fig. 11. photo of the trenching along profile no.11.

400 MHz

 In area 3; Three parallel

AGU Fall Meeting 2015

Using RADAN 7 software and depending on migration procedure on a clear hyperbola from profile no. 1, the dielectric constant value of 12.04 was calculated. This area is closer to the creek and covered by trees.

Area 2:

The fieldwork started on July 21st, 2015, using the SIR-3000 GSSI Inc. equipment with a 400 MHz antenna, a 243 model cart and setting the recording parameters to: 512 samples, 16 bit, 40ns range, 40 scans per unit length.

Fig. 2. Photo showing survey area 1 with location of the 6 profiles; The GSSI SIR-3000 equipped with 400 MHz antenna

Area 1:

Area 3:

Fig. 10. Interpreted GPR data along profile no.9 (Left), photo showing an old rusted key found and verified by digging along the same profile (Right).

Interpreted data of profile no. 9: No clear hyperbolic reflections indicative of the presence of grave sites as no disturbance of the subsurface soil. Animal burrows detected from distance 4.0m to about 12.0m and confirmed on the surface. A small repetitive hyperbolic reflection with high amplitude at distance of 15.0m and depth of about 0.09 m (1.5ns) was verified by digging to be related to an old rusted key.

The Old Carlisle Cemetery Association is interested in searching the area for unmarked or unidentified burial sites and locates any misplaced stones. GPR survey using the 400 MHz antenna was successful in locating many verified burial sites in addition to locate few unmarked graves and misplaced headstones. The association, in particular, requested to know if there are two adjoining graves that are close to burial site no. 17 of area 2 near the east edge of the cemetery. GPR data revealed two clear reflections at depth of about 1.30m and 1.10m and were found to be consistent with the location of a large headstone belongs to burial 17a and 17b. In addition, the north-east corner of the cemetery has been encroached by large trees and private hedge. It is believed that a grave may be outside the cemetery boundary. It is believed there may be unmarked graves in line with burial site no. 10 of area 1. According to family members, this interpretation is as expect, instead of one, at least three burials at this site; two adults and one baby. From GPR data we could see one hyperbolic reflection around a depth of 0.85 m. There are two more hyperbolic reflections around the same location but at a shallower depth of around 0.5 m. This possibly represent the unmarked graves of the two other burials. According to family members, this interpretation is feasible since those two were buried during winter time when the temperature was sub-freezing and they were buried at shallow depths. The head stone of this burial site seems to have shifted at least 1m to the south.

6. Acknowledgements  We thank Mr. Melvin Raborn, representative of the Old Carlisle Cemetery Association Inc., his brother Jerry, and Mr. Bob Elder for their help with the initial logistics.  One author (Hanan Mahdi) thanks the Arkansas Highway and Transportation Department for supporting her during the course of this study.  Taher Hama Ameen thanks University of Sulaimani, Kurdistan Province, Iraq, for supporting him during his visit to UALR.