International Radar Consultants

National Research Institute of Astronomy and Geophysics

Report on

The Geophysical Survey of Tap-Osiris Magna,

Borg El-Arab, Alexandria

Prepared By

Dr. Roger S. Vickers International Radar Consultants

Dr. Abbas Mohamed Abbas National Research Institute of Astronomy and Geophysics

Presented to

The Supreme Council of Antiquities

EGYPT

April 2009

T h e T E A M

EGYPTIAN TEAM

Dr. Abbas Mohamed Abbas Ali

Dr. Hany Salah Abdel-Azim

Dr. Usama Saad Massoud

Dr. Mamdouh Soliman

Dr. Ahmed Kotb El-Emam

Dr. Ahmed Mohsen Lethy

Mr. Abdel-Latif Mohamd Yaunis

Executive Summary

A geophysical survey was performed on the Tap Osiris Magna site near Alexandria, Egypt. The data was concentrated over the Osiris and Isis temples within the site, and confirmed to be of good quality and in the case of GPR showed penetration to 30m. But no remarkable structures were found. We can therefore say that in the areas surveyed, the data shows no major tomb present. There were however, numerous indications of small possible cavities. Using a rule that only those features showing anomalous behavior from two or more independent sensors would be considered, we found only two such regions, both within the temple of Osiris. We therefore recommend that boreholes (3 inch) be drilled in the two areas, with the expectation of connecting with a small cavity or tunnel segment, but not with a major tomb. The data does not, by itself, tell if the structures are man-made or natural, but does give the three-dimensional location of the features that have properties that are different from the background. It is doubtful whether the anomalies detected, by themselves, will result in major find. A third area, indicated by one sensor alone, exists to the East of a known tunnel. For completeness, an additional borehole could be drilled here, but there are no radar echoes to suggest the presence of a tomb or tunnel at this location.

Preamble.

The area to be surveyed was an archeological site approximately 45Km west of Alexandria. It is approximately 88m x 90m and has been excavated (cleaned), resulting in an extremely rough surface (figure 1) which complicates the use of some of the geophysical methods. The geophysical team consisted of a group of scientific staff from the National Research Institute of Astronomy and Geophysics lead by Dr Abbas Mohamed Abbas Ali. The team, together with their areas of responsibility, was;

Dr. Abbas Mohamed Abbas Ali Team leader

Dr. Hany Salah Abdel-Azim ERT

Dr. Usama Saad Massoud VLF/GEM

Dr. Ahmed Kotb El-Emam Mag

Dr. Mamdouh Soliman GPR

Dr. Ahmed Mohsen Lethy Mag

Mr. Abdel-Latif Mohamd Yaunis Assistant

The on site archeologist was Dr. Kathleen Martinez, who worked in coordination with Dr Zahi Hawass , Sec. Gen. of the Supreme Council of antiquities. Dr. Roger S Vickers was retained as an independent consultant primarily for the work with GPR .

Figure 1. Tap Osiris Site looking East showing the rough surface.

The project was conducted to provide information to support or deny the suggestion that the tomb of Cleopatra, and perhaps Mark Anthony lie beneath the Temples of Osiris and Isis inside the TapOsiris Magna complex. The tomb is believed to lie between 20 and 30m below the surface, and is accessed by either a vertical shaft or an inclined tunnel from the surface, possibly originating outside the complex. The local bedrock was limestone/ sandstone of poor quality (many small holes and inclusions) for the first 10m, grading to better quality limestone below that depth.

There were five survey methods used on the site;

Ground Penetrating Radar (GPR)

Electrical Resistivity Tomography (ERT)

Magnetometry (MAG)

Electromagnetic Sounding (EM)

Very Low Frequency Sounding (VLF)

Of these only the first two could be expected to provide information at depths beyond 10m, although all five were capable of providing data in an image format of the topmost few meters in the site. Details of the techniques are given in Appendix I. An overhead view of the site, taken before the recent cleaning work, is taken from Google Earth and is shown in figure 2.

Figure 2. The Tap Osiris Site.

The Survey Work

The team arrived on the site on March 5^th and left on March 13^th. Survey work started at the site at 07:30 and usually finished around 16:30 each day.

The most important task for the first days was to measure the electrical properties of the host rock, so that we could calculate the maximum depth to which the various sensors, particularly the GPR, would be effective. We had been told of ground water table at 33m, and knew that there has been recent rainfall at the site. We were therefore surprised when the first resistivity soundings of the area showed the rock to be quite favorable for GPR work. The rock in fact had a resistivity of 200 ohm meters at depth, which was very good for GPR. Indicating that at low frequencies, the GPR should be able to detect anomalies down to approximately 30m depth.

With that data in hand, the team split up into different units, and proceeded to collect data with all five different sensors. The on-site archeologist stated that the highest probability of finding the tomb lay within the boundaries of the two temples dedicated to Osiris and Isis. The GPR units were therefore used for several days to perform an intense survey of this area. Early on it was found that the higher frequency GPR antennas (100, 200 and 400 MHz) did not perform well over the site, and had very limited penetration ability. The 40 MHz antenna however fulfilled our expectations based on the early measurements of resistivity, and returned echoes from as deep as 30m before becoming noise limited. This antenna was the primary GPR device used for the remainder of the survey, and was deployed over all areas of interest at the site.

Also used over the two temples was the EM and the VLF equipment along with some widely spaced ERT lines. In all, over 140 radar lines were completed, as well as a complete high-density magnetic survey of the entire site, many ERT lines, and EM/VLF data over areas of high interest.

An illustration of the 40 MHz GPR in use is shown in Figure 3,

Figure 3. 40 MHz GPR Antennas in use

The Geophysical Methods used at Tap Osiris Magna

1. GPR

Ground penetration radar is a well known technique for investigating the subsurface in a non-destructive manner. It acts very much like an acoustic echo sounder except that it uses electromagnetic waves instead of acoustic waves. It can be fitted with antennas that radiate different frequencies. Low frequencies, (typically less than 100 MHz) give greater depth of penetration but at low resolution, while higher frequencies (typically 100 - 1000 MHz) give increasingly better resolution at the expense of poorer depth of penetration. Figure 4 shows the 40Mhz antennas being assembled, with the 100 MHz antenna off to the side. The depth ability of GPR is controlled by the resistance of the soils/rocks through which the radar wave has to pass. The higher the resistivity, the better the penetration. Other limitations also affect the GPR's ability, but the resistivity is by far the most important. In many surveys, it is a good idea to take a measurement of this quantity before mobilizing a GPR team.

Figure 4. GPR Antennas being assembled.

2. ERT

Measurements of electrical resistivity have been made for many decades, but more recently, automatic switching units combined with more advanced software have allowed resistivity data to be converted into 2-D cross sections of the subsurface which are much easier to scan visually and to assess the bulk properties of the materials. One disadvantage of the software currently in vogue is that it has a limited color palette and the image shows hard boundaries between regions of increasing/decreasing resistivity, which in turn gives a visual impression of subsurface structure which does not necessarily exist. Nevertheless, it is an important technique, particularly since it is much less subject to the limitations that control GPR performance.

3. EM

The survey here used an instrument called GEM which makes similar measurements to ERT but injects the current into the ground by radiation rather than direct electrodes. The GEM unit transmits six different frequencies, and each one penetrates to a different depth. Typically the maximum depth of penetration is around 6-10m depending as usual on the soil conductivity. The data can be reduced into 2-D depth slices and viewed in a similar manner to GPR slices.

4. Magnetometry

The team used a magnetic gradiometer on the site. This unit measures very small changes in magnetic field. These are caused by the presence of substances with magnetic properties, such as granite clay, or materials containing iron. For example, the presence of granite building blocks in a limestone structure, or clay materials bonding limestone blocks both would give rise to a magnetic signature. The data is usually reduced to a 2-D map, and shows the magnetic properties of the very near-surface. Figure 5 shows the magnetometer in use.

Figure 5. Magnetic Gradiometer in use.

5. VLF.

This technique is passive, i.e. does not contain its own transmitter, and is another method for measuring resistivity . All three resistivity measurement techniques have differing sensitivity to subsurface geometry. One can use different combinations to detect vertical walls, horizontal layers, cavities etc.

Geophysical Data Acquisition

The field procedure at the site of Tap-Osiris Magna temple was started by conducting ERT profiles to inspect the resistivity transition with depth over the studied area. After acquiring the needed information, we planned the best geophysical coverage to achieve the desired goal which was to locate subsurface conditions that indicated the presence of a large tomb, or tunnels that could lead to such a tomb. We were advised by Dr. Martinez that the areas enclosing the tombs of Osiris and Isis would be the most important.

From the early resistivity data we recognized the possibility of using the conducting GPR technique, as the soil resistivity was adequate (increasing with depth) to accomplish the GPR survey to a depth of up to 30m.

The GPR system used was the SIR-3000 from Geophysical Survey Systems, Inc. (GSSI). Figure "6" illustrates the distribution of the GPR profiles that have been carried out over the ground surface of the temple. Some other lines; not showing at this figure; have been conducted around the well-shaft and it will be illustrated later on its own section. The total number of the conducted GPR profiles is 149. Most of them have been measured using the multi-frequency antenna 40MHz. The 100 MHz and 200 MHz antennas were used only at a specific locations (over the tunnel, around the well-shaft and against the southern wall of the well-shaft). All the GPR measurements have been carried out in time-domain. The horizontal distances were measured and then incorporated to the GPR profiles by interpolating the traces along the profile.

Figure 6. Distribution of the GPR profiles over the temple site.

The ERT data were conducted using SYSCAL R2 system from the IRIS company. The system was combined with multi-electrode switching units so automatic switching between the operated arrangement of electrodes could be used to collect data for use in the preparation of 2-D pseudo-sections. . The acquisition was handled utilizing Wenner electrode configuration with equi-offset-distance between electrodes 2.5 or 3 m depending on the maximum available horizontal distance. Sum of 20 ERT profiles were carried out allover the temple site as shown in figure 7.

The electromagnetic survey was performed with two independent procedures, which were: The ground conductivity measurements using GEM 300 device from the GSSI company and the conductivity measurements using the very low frequency "VLF" WADI device from the ABEM company. Those techniques were applied as in figure "8" at 3 grids (A,B and C).

The survey using the magnetic technique was accomplished all over the whole temple site using a gradiometer FM-36.

Figure 7. Distribution of the ERT profiles over the temple site.

Figure 8. Distribution of the GEM/VLF zones over the temple site.

Results

The GPR data was processed to remove the unwanted outside interference signals. Then, time/depth conversion has operated to the whole data set. A sample of the GPR profiles over Osiris Temple are shown through figures (9 and 10), from which it can be seen that interpretation using just individual profiles would be difficult, and so we processed the data to form time (dept) slices over a larger area.

Time-slice processing , converted to depth for the conducted parallel GPR profiles over Osiris temple has been performed and sample of the result is shown on figures 11 (a and b) and 12 (a and b).

In these slices, we looked for echo structure that persisted for more than a single slice,

and also that coincided with high ERT values for the same area.

Figure 9. GPR profile "108"

Figure 10. GPR profile "111"

"a"

"b"

Figure 11. GPR time-slices a) at depth 9.2 m b)at depth 13.8 m.

"a"

"b"

Figure 12. GPR time-slices a) at depth 18.4 m b)at depth 23.0 m.

The ERT data was processed using program RS2DINV to generate the true resistivity tomography section through a least square inversion technique using finite difference method. Examples of the ERT profile results are illustrated through figures (13, 14, 15, 16, 17 and 18). These figures indicate some important features which may have an archaeological meaning. ERT profile "P7" is passing over the tunnel of 8 m depth and it is quite reflected on the profile.

Figure 13. ERT profile P7 passing by part of the tunnel to the south of Osiris temple.

Profile ERT "P8" (figure 14) shows an increase of resistivity with depth and it may indicate some extension for the tunnel to the south of Osiris temple.

Figure 14. ERT profile P9 passing parallel to the eastern edge of Osiris temple.

The anomaly that appears at the eastern part of profile ERT "P15" (figure 15) could be attributed to some extension of the tunnel.

Figure 15. ERT profile P15 passing over the tunnel to the southern edge of Osiris temple.

Three ERT profiles "P18". "P19" and "P20" are passing in part of them through Osiris Temple. Some anomalies were recognized on the profiles and fortunately they are located under Osiris Temple as demonstrated through figures 16, 17 and 18.

Figure 16. ERT profile P18 passing approximately at the middle of Osiris temple.

Figure 17. ERT profile P19 passing at the west of Osiris temple.

Figure 18. ERT profile P20 passing at the western edge of Osiris temple.

The GEM data was collected over 3 areas (A, B and C). The readings of each frequency were contoured using program SURFER 8.0 and exemplified through figures 19, 20 and 21. In general, they can imply the vertical distribution of the conductivity through out the surveyed grids. Some features could be recognized due the boreholes drilled at the site and other limited features, which not implying any archaeological implication.

Figure 19. GEM results of area "A".

Figure 20. GEM results of area "B".

Figure 21. GEM results of area "C".

The VLF survey was completed over the same grids of GEM. The in-phase and quadrature components then contoured to provide the maps which are shown on figures 22, 23 and 24. We can not comment meaningfully on these maps as they do not show any information more than the transition in conductivity. The VLF data set will need considerable extra effort in processing to remove measurement artifacts and extract any interpretive results. At the moment, we do not anticipate any new information from this data set.

Figure 22. VLF results of area "A".

Figure 23. VLF results of area "B".

Figure 24. VLF results of area "C".

Finally, we have to show the resulted map of the magnetic survey (figure 25). It reflects some surface anomalies which could be attributed to fills of shattered potteries and the presence of pieces of some metallic based rocks such as granite. These indications could imply the presence of some caves filled with a collection of shattered potteries or pieces of granite.

Figure 25. Resulted magnetic map over Tap Osiris Temple site.

Conclusions and recommendations

Inspection of the various sensor data sets showed that the data collection was high quality (low noise or interference), and that further repeat collection was not necessary. Examination of the data over a known tunnel (figures 26 and 27) showed that it gave a signature on both radar and resistivity equipment which could then be used if necessary to compare with radar returns in other regions. The radar signature was not as clean as that from resistivity, probably due to edge effects from the two shafts at each end of the test tunnel.

The data so far has shown three areas that are of interest, although none of them have signatures corresponding to a tomb. The areas are ;

1. Resistivity data shows a volume of high resistivity, same values as the known tunnel, to the East of the tunnel and possibly connected with it. There is no corresponding radar signature over the area, and it is therefore the lowest priority of the three we are describing. Nevertheless, a location for a suitable borehole is shown in figure 30.

2. GPR and Resistivity data both show two areas in the Osiris temple which deserve investigation (figures 28 and 29). One of these is in the mid-line of the western edge of the Osiris temple, and the other is at about 20m depth on the SE corner of the temple. We suggest that the fastest and cheapest investigation technique would be to drill three 3 inch boreholes (figure 30) in the site so that either the drill will penetrate the suspected cavity, or if not, will provide the opportunity to put special borehole sensors down the holes and allow the detection of any nearby structures.

Figure 26. The yellow bar indicates the tunnel shaft on ERT profile P15.

Figure 27. ERT profile P17 passes through the tunnel and by Osiris Temple as indicated by the yellow bars.

Figure 28. GPR and ERT anomalies over the SE corner of Osiris Temple.

Figure 29. GPR and ERT profiles at the west of Osiris Temple.

Figure 30. The locations of the suggested boreholes.

January 2, 2021

Roger S.Vickers