J. Barkal Geophysics Report 2023

Report 23005/2023 Gebel Barkal, Sudan Ground Penetrating Radar Survey 2023 For: Geoff Emberling University of Michigan Kelsey Museum of Archaeology 434 South State Street Ann Arbor, Ml 48109 USA By: Eastern Atlas GmbH & Co. KG Berliner Strasse 69 13189 Berlin, Edited: Pawel Wolf Date: 12 July 2024 Contains: 30 pages, 13 maps Germany Eastern Atlas GmbH & Co. KG Jebel Barkal Report 23005/2023 Jebel Barkal / Sudan Ground Penetrating Radar Survey 2023 Report 23005 Page 1/30 Eastern Atlas GmbH & Co. KG Jebel Barkal Table of Contents 1. Objectives and investigated areas ....................................................................... 5 1.1 Introduction .............................................................................................................. 5 1.2 lnvestigated areas ................................................................................................... 10 2. Methodology, Technical Parameters, and Data Processing ............................ 11 2.1. Principles of GPR Prospection................................................................................... 11 2.2 The Pulse EKKO System ............................................................................................ 12 2.3 The GSSl System ....................................................................................................... 13 2.4 Topographic Measurements..................................................................................... 14 2.5 Terrain documentation by UAV overflight ................................................................ 15 3. Archaeological Interpretation ............................................................................ 16 3.1 Preparation of site maps and other resources .......................................................... 16 3.2. Presentation of the GPR data ................................................................................... 16 3.3. Notes on the archaeological interpretation of the data ............................................ 17 3.4 Description and archaeological interpretation of the GPR areas ............................. 18 3.4.1 Survey area JBG ................................................................................... 18 3.4.1.1 General description .............................................................................. 18 3.4.1.2 Archaeological interpretation ............................................................... 18 3.4.2 Survey areas JBF/JBH/JBI..................................................................... 20 3.4.2.1 General description .............................................................................. 20 3.4.2.2 Archaeological interpretation ............................................................... 20 3.4.3 Survey area JBJ .................................................................................... 25 3.4.3.1 General description .............................................................................. 25 3.4.3.2 Archaeological interpretation ............................................................... 26 3.4.4 Survey areas JBK/JBL ........................................................................... 26 3.4.4.1 General description .............................................................................. 26 3.4.4.2 Archaeological interpretation ................................................................... 3.5 Recommendations for future GPR prospection ........................................................ 28 4. Summary ............................................................................................................ 30 Report 23005 Page 2/30 Eastern Atlas GmbH & Co. KG Jebel Barkal Table of Figures Fig. 1 Photograph of the Reisner excavations around temple B500 ....................................... 5 Fig. 2 Areas covered by magnetic gradiometry and excavations in seasons 2006-2007........... 6 Fig. 3 Part of the excavations at the Eastern Mound in 2023 ................................................. 7 Fig. 4 Areas JBC/JBD of the GPR survey in 2022 with indication of urban structures ............... 8 Fig. 5 J. Garstang’s preliminary plan of the northern part of the Royal City of Meroe ............ 9 Fig. 6 Original proposal of the GPR survey areas for season 2023 .......................................... 9 Fig. 7 Geophysical survey areas at J. Barkal since 2006......................................................... 11 Fig. 8 The Pulse EKKO System with two SPIDAR 250-MHz antennas ..................................... 12 Fig. 9 GSSI system SIR 3000 with 400-MHz-antenna ............................................................. 13 Fig. 10 The NovAtel GNSS receiver used as rover ................................................................... 14 Fig. 11 Section south in trench 23-05 .................................................................................... 18 Fig. 12 Intrusive wadi khor in the NW part of area JBF/JBH (ortho-photographic image) ....... 21 Fig. 13 Intrusive wadi khor in the NW part of area JBF/JBH (DEM with relief filter) ................ 21 Fig. 14 Area JBF/JBH during the survey seen from south ........................................................ 22 Fig. 15 Results of former magnetic gradiometry surveys in the area of JBF/JBH and JBJ ......... 23 Fig. 16 Field drawing of the trenches excavated 2006 at the northwestern side of B1200 ..... 25 Fig. 17 Temples KC 100 and M720 in Meroe City ................................................................... 27 Fig. 18 Proposal for future geophysical prospection .............................................................. 29 Index of Tables Table 1 Coverage, location and brief description of the surveyed areas ............................... 10 Table 2 Technical parameters of the Pulse EKKO System ..................................................... 12 Table 3 Technical parameters of the GSSI System ................................................................ 13 Table 4 Technical parameters of the DJI Mavic 2 Pro drone ................................................. 15 Report 23005 Page 3/30 Eastern Atlas GmbH & Co. KG Jebel Barkal Maps in the Appendix Map No Content 23005-1 23005-1.1 23005-1.2 Overview maps Location of GPR survey areas 2020-2023 Former research: GPR survey areas JBF – JBL in relation to former excavations and prospection: Reisner's excavation, magnetic gradiometry 2006ff, GPR surveys 2020/2022 with interpretation Recent and natural features (with depth slices for time interval 0 - 4 ns) Archaeological interpretation of GPR surveys 2020-2023 GPR survey areas JBA - JBL: Summarised depth slices for time interval 0-4 ns GPR survey areas JBA - JBL: Summarised depth slices for time interval 6-22 ns GPR survey areas JBA - JBL: Summarised depth slices for time interval 24-38 ns 23005-1.3 23005-1.4 23005-1.5 23005-1.6 23005-1.7 23005-2 23005-2.1 a-c 23005-2.2 d-f 23005-3 23005-3.1 a-c 23005-3.2 d-f Scale GPR survey area JBG (a) Former excavations and prospection; (b) Recent and natural features; (c) Archaeological interpretation Summarised depth slices: (d) 0-4 ns; (e) 6-22 ns; (f) 24-38 ns GPR survey areas JBF/JBH/JBI (a) Former excavations and prospection; (b) Recent and natural features; (c) Archaeological interpretation Summarised depth slices: (d) 0-4 ns; (e) 6-22 ns; (f) 24-38 ns 23005-4 23005-4 a-f GPR survey area JBJ (a) Former excavations and prospection; (b) Recent and natural features; (c) Archaeological interpretation; Summarised depth slices (d) 0-4 ns; (e) 6-22 ns; (f) 24-38 ns 23005-5 23005-5 a-f GPR survey areas JBK/JBL (a) Former excavations and prospection; (b) Recent and natural features; (c) Archaeological interpretation; Summarised depth slices (d) 0-4 ns; (e) 6-22 ns; (f) 24-38 ns Report 23005 1:1500 1:1500 1:1500 1:1500 1:1500 1:1500 1:1500 1:500 1:500 1:1200 1:1200 1:750 1:750 Page 4/30 Eastern Atlas GmbH & Co. KG 1. Objectives and investigated areas 1.1 Introduction Jebel Barkal The report presents the results of a ground-penetrating radar (GPR) survey conducted between January 9 and 21, 2023, at selected areas assumed to represent parts of the urban royal settlement of Napata at Jebel Barkal (Karima, Northern State, Sudan). The survey was carried out by Eastern Atlas/Berlin1 on behalf of the Jebel Barkal Archaeological Project (JBAP)2, directed by G. Emberling and Sami Elamin, and the company VisualSkies3. Napata, the Kushite major sacred and administrative center in ancient Nubia, has been assumed to be located in the area in front of Jebel Barkal that was occupied by monumental sacred temples since the Egyptian New Kingdom occupation of Nubia. The site is well-known for its Egyptian and Kushite temples, shrines, and large palatial monuments, which indicate its significance as a main religious and administrative complex of the Kushite kingdom. During 1916 to 1920, the Harvard University - Boston Museum of Fine Arts Expedition, directed by G. A. Reisner, initiated the first systematic large-scale excavations in front of J. Barkal. Conducted in order to uncover the remains of the ancient town of Napata, the excavations focused on the Amun Temple B500 and surrounding areas, where a number of further temples, shrines, and palaces were unearthed (Fig. 1). While these findings provided a first comprehensive view of the site’s sacral and political significance, no tangible evidence of a substantial urban settlement, such as dense domestic structures, streets and public places, had been uncovered. Fig. 1 Photograph of the Reisner excavations around temple B500 (© MFA Boston, Neg. No. A2928_NS) 1 The team members were B. Ullrich, R. Knieß, M. Tielmann, and P. Wolf. University of Michigan, Kelsey Museum of Archaeology, 434 South State Street, Ann Arbor, MI 48109 USA. 3 VisualSkies, Unit 9S Hewlett House, 5 Havelock Terrace, Battersea, London SW8 4AS, UK. 2 Report 23005 Page 5/30 Eastern Atlas GmbH & Co. KG Jebel Barkal Building upon these initial excavations, a mission from the University of Rome La Sapienza, initially led by S. Donadoni and subsequently by A. Roccati, engaged in further excavations of the temples and palaces since 1973. Its principal discovery was the large palace B1500, constructed during the reign of the Meroitic king Natakamani. Under the present direction of E. Ciampini, the mission focused its efforts on the Meroitic "Natakamani District" to the northeast of temple B500 exploring further Meroitic palatial structures. Since 1986, T. Kendall (originally under the sponsorship of the Museum of Fine Arts, Boston, later under the umbrella of NCAM, Khartoum, and the Northeastern University, Boston) led a long series of excavation seasons, with particular focus on the Amun temple B500 and surrounding Egyptian and Napatan-Meroitic shrines, including excavations at e.g. the Anlamani-Aspelta palace B1200 and other significant spots. In 2000, his team conducted initial geophysical surveys directed by M. S. Watters using magnetic gradiometry and GPR northeast of temple B500 (discovering parts of B1700) and in the area of B1150. Expanded magnetic gradiometry surveys were conducted in 2006 and 2007 by T. Goldmann, R. Wutzler, and P. Wolf, along with test trenches in various parts of the survey area dug by P. Wolf and U. Nowotnick (Fig. 2). These magnetic gradiometry surveys detected new temples such as B560, B561 and B570 in front of the Amun temple B500. The Meroitic double-structure B560/561 was thereafter excavated in 2014-2015 by J. Knudstad, R. Frey, J. Haynes, and M. Osman, and the Meroitic structure B1700, probably a priests’ house to the northeast of temple B500, was excavated since 2015 by M. Lebedev and S. Nannucci. Fig. 2 Areas covered by magnetic gradiometry and excavations in seasons 2006-2007 (© P. Wolf, 2007) At present, the Jebel Barkal Archaeological Project (JBAP), co-directed by G. Emberling and El-Hassan A. Mohamed, later by Sami Elamin, employs advanced geophysical techniques and large-scale excavations to investigate the broader historical context of Napata. The objective of this project is to gain insight into the urban layout of Napata and its socio-political structure. Since 2016, G. Tucker conducted several large-scale magnetic gradiometry surveys in several areas at the site, including the Report 23005 Page 6/30 Eastern Atlas GmbH & Co. KG Jebel Barkal so-called "Eastern Mound."4 Here these surveys identified subsurface features including settlement patterns suggesting a complex urban layout including dense building structures and potential street networks. Additionally, GPR surveys conducted in 2020 by B. Ullrich, R. Kniess, and G. Tucker revealed various near-surface building remains in this area. Based on these results, excavations conducted by T. Skuldbol revealed multi-level building structures and confirmed the existence of a large street lined with buildings on either side at the Eastern Mound (Fig. 3). These findings indicate significant domestic activities dated to the 1st century BC / 1st century AD and confirm that during the Meroitic period, Napata featured not only a religious and administrative center but also a densely populated urban area. Fig. 3 Part of the excavations at the Eastern Mound in 2023; drone photograph taken by S. Elamin However, despite all these archaeological efforts, the exact location and extent of an originally Napatan urban settlement remained unclear. Neither geophysical prospection nor archaeological excavations revealed clear domestic structures or a comprehensive street network dating to the Napatan period, which spanned from the 9th to the 4th century BC, prior to the Meroitic period of the Kushite kingdom. While it was still assumed that significant communities must have lived and worked here during the Napatan period, clear archaeological evidence of a Napatan urban settlement was still missing. In 2022, a GPR survey conducted by B. Ullrich and P. Wolf focused on identifying new archaeological features around temple B500 and farther south between palaces B1200 and B100 (hereafter referred to as the Barkal Palace Area, BPA). In addition to providing new insights into the known structures of 4 Tucker, G and G. Emberling, ‘Settlement in the Heartland of Napatan Kush: Preliminary Results of Magnetic Gradiometry at El-Kurru, Jebel Barkal and Sanam’, Sudan & Nubia 20 (2016), 50-57. Report 23005 Page 7/30 Eastern Atlas GmbH & Co. KG Jebel Barkal palace B1200, including the discovery of a large open courtyard enclosed by a substantial adobe wall, the survey identified several trapezoidal buildings divided by approximately 2.5m wide streets adjacent to the newly identified courtyard of B1200 (Fig. 4). These findings suggested the presence of complex urban structures in the previously unexcavated area between the two palaces – with buildings that are likely domestic villas and streets running more or less orthogonally between them. Their internal structures and the overall layout suggested a well-organized urban pattern similar to that observed in other Meroitic sites such as Meroe, Hamadab, and Kawa (Fig. 5). They suggested that the BPA was densely populated with residential and administrative buildings being part of a settlement with clearly urban character and challenging the previous notion of large open spaces around the monumental sacral and palatial structures at the site. This new understanding aligns with the idea that the site was not just a religious center but also an important urban hub already during the Napatan period. As these findings underscored the necessity for additional GPR surveys and targeted excavations to more comprehensively examine the Napatan urban settlement at Jebel Barkal, the opportunity was seized to pursue the GPR survey in 2023, with funding from the National Geographic documentary "Lost Cities Revealed with Albert Lin." Fig. 4 Areas JBC/JBD of the GPR survey in 2022 with indication of the urban structures in the BPA (marked in red; cf. GPR report 2022) Report 23005 Page 8/30 Eastern Atlas GmbH & Co. KG Jebel Barkal Fig. 5 J. Garstang’s preliminary plan of the northern part of the Royal City of Meroe5 Fig. 6 Original proposal of the GPR survey areas for season 2023 (proposed areas marked yellow) 5 After Hinkel, F.W. and U. Sievertsen, 2002, Die Royal City von Meroe und die repräsentative Profanarchitektur in Kusch, in: F.W. Hinkel (ed.), The Archaeological Map of the Sudan Supplement IV, Berlin 2002, pl. IX.8). Report 23005 Page 9/30 Eastern Atlas GmbH & Co. KG 1.2 Jebel Barkal lnvestigated areas Based on the results of the geophysical surveys since 2006, we assumed that a large number of dense and still unknown Napatan domestic and other urban settlement structures are located on both sides of the Amun temple B500 in front of J. Barkal. In particular, northeast of the Amun temple B500 and west of the ‘Natakamani District’, several large rectangular anomalies had been prospected by the magnetic gradiometry surveys in 2006-2007 indicating the presence of temple-like and palatial monuments (cf. Fig. 2). Some of these structures had been later-on excavated by the Italian Mission and turned out to be representative palace and villa-like monuments as well as sacred shrines. However, infrastructural relations between them such as roads and pathways as well as public open spaces had never been tested. Our original plan was therefore to survey several contiguous areas northeast and southwest of temple B500 in order to cover the entire territory (Fig. 6). The main focus was put on the area northeast of the temple, as it is less impacted by Reisner's excavation dumps.6 However, since the GPR team did not get permission to work in the Italian Mission’s license area, which covers almost the entire terrain northeast of the B500, the plan had to be changed and we switched to smaller areas between the excavation dumps of Reisner southwest of the B500 temple. Figure 7 and Maps 23005-1.1 to 1.3 in the attachment illustrate the 2023 surveyed areas designated JBG, JBF/JBH, JBI, JBJ, JBK, and JBL in conjunction with the areas surveyed by magnetic gradiometry and GPR in previous seasons. Table 1 provides a brief descriptive overview of these areas. Their detailed descriptions can be found in chapter 4.3 along with the presentation of the GPR results. Before the survey, all areas were roughly cleared from larger debris and larger grass bushes were cut back by a 10-man strong group of local workmen. Designation Covered area JBF/JBH 9097 m2 JBG 2348 m2 JBI 1232 m2 JBJ 2531 m2 JBK 2496 m2 JBL 950 m2 Location Brief description next to fence at SW limit of the site, spanning from B300 to NW of the pathways in the SE of the site flat & horizontal area, few surface disturbances by smaller excavation dumps and a recent pottery kom, western part disturbed by fluvial erosion windblown sand dune with larger between two large excavation dumps SW of forecourt B501 and area in front of bushes in the eastern part B500 at SE limit of the site, abuts JBF/JBH and flat area covered by many bushes JBC at their eastern side and grass outcrops, partially disturbed by modern car tracks NW of palace B1200, abuts JBF/JBH heavily disturbed by water gullies, excavation dumps and debris of B1150 SE half of the processional area in front depression, just superficially of B500 and NW of modern main track; disturbed by modern car tracks and covers B570 and processional way few medium-size bushes abuts JBK to the NE sloping towards JBK, disturbed by excavation of B561 and larger bushes Table 1 Coverage, location and brief description of the surveyed areas 6 As has been shown by the GPR tests in 2020, GPR can penetrate these up to several meters high dumps using lowerfrequency waves, but the object resolution then would probably suffer. Report 23005 Page 10/30 Eastern Atlas GmbH & Co. KG Jebel Barkal Fig. 7 Geophysical survey areas at J. Barkal since 2006, including the areas surveyed by GPR in 2023 (marked in yellow) 2. Methodology, Technical Parameters, and Data Processing 2.1. Principles of GPR Prospection The GPR (ground-penetrating radar) method is based on the propagation of high-frequency electromagnetic waves into the ground. These waves are reflected and refracted by different layers and objects, such as stones and walls. The registration of transit time differences and amplitudes of the electromagnetic waves provides information about the position, depth, and specific properties of objects and layers buried in the ground. The spherical resolution and depth of penetration depend on the GPR antenna’s frequency and the electromagnetic properties of the ground. A general rule of thumb is that the higher the frequency, the better the spherical resolution, but the lower the depth of penetration. The propagation conditions of electromagnetic waves are mostly determined by the soil properties. The main factor is the ground moisture, as water has a very high dielectric permittivity ε, which causes a strong attenuation of electromagnetic waves. For this reason, dry ground offers more favorable conditions than saturated soils. Another important influence comes from clay minerals. In many instances, the depth of penetration and resolution of GPR measurements in clayey soils are limited. This is attributed to the presence of crystallized water that is bound with clay minerals.7 7 For more information on GPR see Jol, H. M., Ground Penetrating Radar: Theory and Applications, Amsterdam/Oxford: Elsevier Science, 2009. Report 23005 Page 11/30 Eastern Atlas GmbH & Co. KG 2.2 Jebel Barkal The Pulse EKKO System In season 2023 the team used two different sets of technical equipment, which both had been tested at the site in seasons 2020 and 2022. In areas JBF/JBH, JBI, JBJ, JBK, and JBL, the Pulse EKKO system SPIDAR with 250-MHz antennas was utilized (Fig. 8). The highly adaptable system incorporates data recording devices and two antenna pairs with distinct receiver and transmitter antennas. The antennas were mounted in parallel, enabling measurements to be conducted as a two-channel system to more effectively survey larger areas. The cart with the antennas was employed for measurements in Common Offset Mode to cover the survey areas by single or parallel profiles. For the positioning of data, a RTK-GNSS and a survey wheel for profile length measurements were mounted on the SPIDAR. For data processing the software REFLEXW (Sandmeier Scientific, Germany) was used. Table 2 presents the principal technical data of the system. Method Ground Penetrating Radar Survey System Pulse EKKO Sensors SPIDAR 250-MHz antennas Measurement Category Configuration Two-way transit time and amplitude of reflected electromagnetic waves Survey resolution 0.5 m profile distance, 0.05 m point distance Positioning RTK-GNSS; Additionally Odometer Data Processing REFLEXW (Sandmeier Scientific, Germany) Data Formats Raw data: DZT, processed data: REFLEXW format, time slices: ASCII Two parallel antennas with survey wheel Table 2: Technical parameters of the Pulse EKKO System Fig. 8 The Pulse EKKO System with two SPIDAR 250-MHz antennas Report 23005 Page 12/30 Eastern Atlas GmbH & Co. KG 2.3 Jebel Barkal The GSSl System In area JBG that was partially covering a windblown sand dune between two large excavation dumps of Reisner, a SlR-3000 system from GSSl (USA) with a shielded 400-MHz antenna was used (Fig. 9). The antenna was switched to transmitter and receiver mode and therefore was able to detect structures directly below the antenna as well as in greater depth. For the positioning of data, the corner points of area JBG were surveyed by differential GPS (see below chapter 2.4), combined with a survey wheel for profile length measurements. Data processing using the software REFLEXW included a migration to avoid an overinterpretation of hyperbolas. A topographical correction was not added, as most of the area in JBG was flat and horizontal. Table 3 presents the principal technical data of the system. Method Ground Penetrating Radar Survey System SlR 3000 (GSSl) Sensors 400-MHz-antenna, model: 5040 Measurement Category Two-way transit time and amplitude of reflected electromagnetic waves Configuration One antenna with survey wheel Survey resolution 0.5 m profile distance, 0.05 m point distance Positioning Odometer for profile records; RTK-GNSS for corner of survey area Data Processing REFLEXW (Sandmeier Scientific, Germany) Data Formats Raw data: DZT, processed data: REFLEXW format, time slices: ASCll Table 3: Technical parameters of the GSSI System Fig. 9 GSSI system SIR 3000 with 400-MHz-antenna Report 23005 Page 13/30 Eastern Atlas GmbH & Co. KG 2.4 Jebel Barkal Topographic Measurements The data positioning for the GPR survey in area JBG and for additional points was realized by means of a differential GPS using two NovAtel GNSS receivers as base and rover (Fig. 10). The relative accuracy in RTK mode was mostly ±2 cm. The GNSS base antenna was mounted to the roof of the JBAP mission’s dig house in Karima and its position was post-processed using RlNEX data for the calculation of baseline to lGS station Addis Abeba. The GNSS rover antenna was used to survey the corners of the JBG area that was laid out by tapes and lines, and to survey the four main station points (set in concrete) of the JBAP excavations at the Eastern Mound. Four additional station points were set out for the test excavations at the BPA that were carried out after the GPR survey by P. Wolf. All surveyed data was referenced to WGS84 / UTM Zone 36 (EPSG 32636). Fig. 10 The NovAtel GNSS receiver used as rover Report 23005 Page 14/30 Eastern Atlas GmbH & Co. KG 2.5 Jebel Barkal Terrain documentation by UAV overflight To improve the ortho-photographic background maps and the digital elevation model (DEM) of the GPR surveyed areas, the existing images of the terrain in front of J. Barkal were updated with new UAVborn photos taken immediately after the GPR survey and featuring a better resolution than the previous images. A DJI Mavic 2 Pro drone was used for this purpose. Since the positioning system of the drone is not accurate enough for the calculation of 3D-models and ortho-photographic images with an accuracy in the cm-range, it was not used for this purpose. Instead, a set of set out and ‘naturally present’ ground control points were utilized: Apart from the four station points established by the team in the BPA, prominent points on various ground structures and monuments (e.g. stone block corners; markers such as cracks or carvings on stone blocks; concrete posts etc.) were used to align and reference the new photos to the existing ortho-photographic images that were produced already in 2022 and which exhibited a sufficient accuracy. Table 4 presents the principal technical data of the system as well as flight attitude and ground resolution of the raw photos. Method AUV Photography System DJI Mavic 2 Pro Camera Hasselblad L1D-20c Photo sensor Lens 1-Zoll CMOS, 20 Mpix 28mm f/2.8-11, FOV 77° Positioning system GPS and GLONASS Positioning accuracy horizontally ± 1.5 m, vertically ± 0.5 m Flight altitude 10-15m Ground resolution 5 mm / pix Table 4: Technical parameters of the DJI Mavic 2 Pro drone Report 23005 Page 15/30 Eastern Atlas GmbH & Co. KG 3. Archaeological Interpretation 3.1 Preparation of site maps and other resources Jebel Barkal In order to correctly interpret the GPR data and incorporate it into the site plan, the site plan that was in use until 2022 had been corrected and referenced using the re-surveyed main station points at the Eastern Mound as well as other ground control points already in season 2022:8 For that purpose, in summer 2022 a new site plan was prepared on the basis of a high resolution ortho-photograph of the terrain in front of the J. Barkal that was calculated on the basis of a 3D model of the area using c. 2700 drone-based photographs taken by S. Elamin during overflights in November and December 2021. In order to enhance the surface interpretation of the areas surveyed by GPR in the present season of January 2023, all GPR areas have been overflown again shortly after the conducted GPR survey and rephotographed by P. Wolf in order to obtain a new ortho-photographical basemap with a better ground resolution. About 4600 drone photographs were taken during these overflights and used to recalculate the 3D-model and the ortho-photographical basemap of the area in front of J. Barkal. The mapping of vegetation and other recent surface features such as tracks and pathways, excavation areas and dumps, which was undertaken in 2022 to take into account disturbing non-archaeological objects and recent structures in the interpretation of the GPR data, was extended to the areas surveyed in 2023. For this purpose, various published plans, photos of Reisner's excavations and the results of the former magnetic gradiometry surveys have been used. 3.2. Presentation of the GPR data The results of the GPR surveys are presented in this report as a set of [1] overview maps of all investigated areas (maps 23005-1.1 to 23005-1.7), as well as [2] in separate maps of the areas JBG, JBF/JBH, JBI, JBJ, JBK, and JBL (maps 23005-2 to 23005-5). For [2], the sub-maps (a) of each map set illustrate the corresponding GPR area with indication of previously excavated areas and magnetic gradiometry data where applicable. (b) illustrates recent and other non-archaeological faults, drawn overlying an image of the summarized GPR depth slices of 0-4ns9: Vegetation [dotted green], pathways and tracks [brown], previously excavated areas [hatched], test trenches [gray] and dumps [dotted gray], areas of increased ground moisture [blue], as well as larger stones and stone blocks [yellow], and other recent stuff [purple]. They were determined from the ortho-photographic base map and the datasets mentioned above. Areas with high ground moisture were determined based on morphological features such as drainage gullies and channels, as well as through interpretation of the GPR data. (c) presents the archaeological interpretation with line drawings on the summarized GPR depth slices of 6-22ns (cf. chapter 3.3): The lines indicate walls, blocks, trenches, and similar structural features. Their color indicates the assumed material of the features such as stone walls and blocks (ochre), adobe walls (dark red), jalous walls (brown). It should be noted, however, that unless verified by excavation, which is most rarely the 8 9 Using Datum: WGS 84 / UTM zone 36N [EPSG 32636]. For the interpretation of the actual physical depth of a slice related to a corresponding time interval see chapter 3.3. Report 23005 Page 16/30 Eastern Atlas GmbH & Co. KG Jebel Barkal case, these color designations are assumptions, most often based on the context and shape of a feature. Low radar amplitudes marked by a yellow hatched pattern are interpreted as debris, which possibly accumulated after the destruction or during the decay of the corresponding structures. Individual blocks and boulders were included in this map as well, as it is a priori unclear whether they are to be counted as in situ archaeological objects or whether they were here in a secondary position. In order to match the existing site plans and the plans of the test excavations in the BPA (carried out directly after the GPR survey), different colors were used in the following two cases: (black) if the corresponding features were already known10, and (green) if they were verified by the test excavations. The line type refers to the reliability of the interpretation: solid lines denote clearly recognizable features, dashed lines indicate features not clearly recognizable, while dotted lines stand for those cases where the assumption predominates. Dashed black lines outside the GPR squares mark wall structures that are known from earlier plans but are not visible on the ground surface today (this distinction from visible walls was, however, not consistently followed in the maps here). Finally, small letters (enclosed in circles) designate certain feature groups, which are referred to in the descriptions and interpretations in chapter 3.4. (d) – (f) illustrate the GPR data in summarized depth slices for the depths of 0-4ns (d), 6-22ns (e) and 24-38ns (f)11. They serve mainly to control the archeological interpretation by giving the reader the chance of an independent own interpretation. 3.3. Notes on the archaeological interpretation of the data The archaeological interpretation is based on the delineation of distinct areas characterized by significant variations in reflection amplitudes, both in terms of amplitude and shape, as well as lateral and vertical changes in these amplitudes. The datasets collected in the field along parallel survey lines were processed using the software REFLEXW (Sandmeier Scientific, Germany) to be presented in horizontal slices. These slices display the reflection amplitudes of time intervals up to 38ns in grayscale, where white represents low electromagnetic wave amplitudes and black represents high amplitudes. For practical purposes, these slices have been clustered into three groups: 0-4ns, 6-22ns, and 24-38ns, as it was found that many subsurface structures are more discernible when viewed in this manner, given that the structures are typically three-dimensional and thus have depth. The time intervals correlate with depth intervals that were calculated at an average electromagnetic wave velocity of 0.15 m/ns. This velocity was derived from the CMP measurements taken on Reisner’s excavation dump in area A4/5 during the 2020 survey. Based on this velocity, the expected depth intervals are: - 0-4ns = 0-0.75m - 6-22ns = 0.9-3.3m - 24-38ns = 3.6-5.7m - 40-48ns = 6-7.2m. After the GPR survey, a test excavation in a 20x20m area within the GPR area JBC (surveyed in 2022) was conducted by P. Wolf, with assistance from K. Rose and J. Sanches, to validate the GPR results. The excavations confirmed that the measured electromagnetic wave velocity of 0.15 m/ns is accurate: within the deepest test trench BP23-05, virgin ground was reached at a depth of 2.3m (Fig. 11). 10 11 E.g. because they were visible on ground or had been excavated by previous excavations. See chapter 3.3 for a more detailed explanation of the time interval / reflection depth relation. Report 23005 Page 17/30 Eastern Atlas GmbH & Co. KG Jebel Barkal Therefore, it is assumed that human-made structures such as walls will be most frequently observed in the GPR time slices of the two shorter intervals up to 22 ns, corresponding to depths of 0-3.3 m below the present surface. This assumption was corroborated by the interpretation of the GPR data: - depth slices of up to 4ns time interval (down to 0.75m) clearly show surface features such as vegetation, roots, pathways, and larger structural remains like walls, - medium depth slices of 6-22ns time intervals (corresponding to 0.9-3.3m) reveal many larger-scale and orthogonal structures, which can be interpreted as ancient walls and similar structures more clearly than in the previous time interval, - lower depth slices between 24-38ns (i.e., 3.6-5.7m) primarily show diffuse, likely geological morphologies, along with many reflections of near-surface structures, - depth slices with the time intervals of 40-48ns corresponding to depths of 6-7.2m were not used for the archaeological interpretation. The division of the data into the first three summarized depth intervals aids in the interpretation of the GPR data. Since the grouping depends on the physical volume of the subsurface objects and their physical ‘depth’ as well as their depth in the ground, the decision on the number of groups and the size of the time intervals is based on experience and will vary from site to site. In this context, an interesting paradox arises. Typically, objects with higher moisture content reflect waves with higher amplitudes, appearing darker in the slices than drier objects. However, in some cases, structures that can be clearly interpreted as walls appear light in the middle and lower depth slices (i.e., with low amplitudes) but exhibit higher amplitudes (darker) in the upper depth slices. This 'paradox' could result from the physical properties of the adobe walls, particularly regarding subsurface moisture at various ground levels. However, it might also be due to the fact that ancient walls were often excavated by farmers to use the muddy material for fertilizing their fields (marogh digging). Since these holes and trenches were typically not backfilled by the farmers but gradually silted up, their higher sand content (which retains moisture better than clayey material) could be responsible for the higher amplitudes and thus the darker appearance of the corresponding structures. Fig. 11 Section south in trench 23-05 indicating the virgin ground at 2.3m depth in this spot Report 23005 Page 18/30 Eastern Atlas GmbH & Co. KG 3.4 Jebel Barkal Description and archaeological interpretation of the GPR areas 3.4.1 Survey area JBG (Maps 23005-2.1 and 2.2) 3.4.1.1 General description Area JBG, which covers an essentially rectangular area of 30x100m (2348 m2), runs almost north-south, with its northern edge roughly abutting the southwestern wall of forecourt B501, while its southern edge lies in front of a patch of larger grass shrubs near the modern main track, which borders the BPA to the southeast. Due to its sandwich-like position between two of Reisner's large excavation dumps, it runs on a 1.5-2 m thick dune of wind-blown sand that has accumulated in this low-wind depression. This is probably why the ground moisture here is higher than elsewhere, because the extensive sand body retains moisture longer than other ground materials. Not only does JBG differ in these respects from all other GPR areas which have been surveyed this season, it is also the only area where the SIR 3000 system has been used. Therefore, the GPR results may differ slightly from those of the other areas. The northern third of area JBG overlaps a large trench of Reisner's excavation of temple forecourt B501. Apart from this, no other excavations or magnetic gradiometry surveys have yet been carried out in this area. 3.4.1.2 Archaeological interpretation Due to the somewhat challenging conditions in area JBG, we did not expect much here, particularly because of the heavy sand coverage and the expected high soil moisture. Nevertheless, the GPR detected some seemingly human-made structures, whose interpretation is, however, associated with some uncertainties due to the relatively small survey window. In general, however, the GPR results make us optimistic that also Reisner’s larger spoil heaps can be surveyed with relatively good results using a lower frequency antenna as on the Pulse Ecco with 250 MHz. Next to the southwestern outer wall of courtyard B501, the GPR shows some apparently larger structures (a), some of which have the same orientation as B501, while others have a slightly different orientation. Given their size and orientation, they may be reminiscent of the foundation remains of earlier sacred buildings, unless this is just a misinterpretation of the GPR reflections, and these reflections originate in fact from squatter occupations as not all features that Reisner documented in this area could be included in the GPR maps - especially not the many squatter occupation walls inside and outside forecourt B501 (cf. Fig. 1). The reflections are located in the area of Reisner’s large trench, and it is quite probable that Reisner had removed the squatter occupations during the process of excavation.12 The interpretation is additionally supported by the fact that the structures are best recognizable in the lower depth slices. Some meters farther southeast, partially overlapping the features mentioned under (a), a group of round features (b) are similarly best recognizable in the deepest time slices. Their diameters vary between 2m and 10m, mostly around 5m. Their reflections have high amplitudes and therefore may originate from moisture in the ground. It seems that a line of medium-sized reflections runs in a distance of c. 15m parallel to the outer wall of B501 towards northwest. Extending this line further into the GPR area JBB of season 2022, one encounters a row of 4 similar reflections. Compared to other Kushite sites such as Musawwarat, Hamadab and Meroe, it would be quite plausible to interpret these reflections as originating from planting pits. In the Great Enclosure at Musawwarat, for example, the plantation pits of the temple garden in courtyard 117 run in two parallel 12 Even if not, it wouldn’t be clear whether these walls would have survived to this day. Report 23005 Page 19/30 Eastern Atlas GmbH & Co. KG Jebel Barkal lines at a distance of 10-15m along the façade of the terrace of the main temple IA 100.13 Plantation pits with diameters of up to 5m have been discovered by J. Garstang in Meroe City, where they line up the northern main avenue leading into the Royal City (cf. Fig. 5). The remains of a brick construction, reminiscent of a Meroitic water conduit, stick few meters further northwest – roughly at the level of the second pylon – out of the ground (c). Extending the line of medium-sized reflections farther northwest, it would meet the large well B1000 near the rear part of the temple (cf. Fig. 1, foreground). The plantation pits documented in Musawwarat and in Hamadab are linked by open and partially covered water channels with wells (Hamadab) and cisterns (Musawwarat).14 If these reflections indeed represent plantation pits of a temple garden, several pits excavated by Reisner to the northeast of temple B500 (GPR area JBA of 2022) may likewise represent plantation pits. A few meters southeast of the group of these reflections (b), are some linear reflections (d) that are, however, too weak to be meaningfully interpreted with any certainty. More clearly visible are larger structures (e) in the southeasternmost quarter of JBG. These could originate from a larger building the orientation of which seems to roughly align with that of temple B570 and the presumed processional avenue (see below chapter 3.4.4.2). The thickness of its walls likewise suggests a sacred building that was erected in a 2nd row along the processional avenue.15 3.4.2 Survey areas JBF/JBH/JBI (Maps 23005-3.1 and 3.2) 3.4.2.1 General description As these GPR areas adjoin each other and partly overlap, they are described jointly in this chapter. Area JBF/JBH is a strip of approximately 200x40 meters (9097 m2), oriented in a northwest-southeast direction along the southwestern boundary of the J. Barkal archaeological site. It extends from the smaller spoil heaps near the temples B200 and B300 in the northwest to nearly the grass tufts and tracks that separate the archaeological site from the main track near the palm plantations in the southeast. Area JBI forms an elongated rectangle of c. 15x82m (1232 m2) that abuts and partially overlaps the southeastern end of JBF/JBH.16 The terrain in this area is relatively flat and slopes gently to the southeast. Situated in the shelter of J. Barkal, it is free from windblown sand dunes. Its ground consists of a mixture of fluvial deposits and weathered building remains – predominantly gravel and silt with smaller stones, pebbles and some clay content. The alluvial deposits seem to have been washed-in mainly by a wadi that flows in from the west into the area opposite temples B200/B300. Its fan-shaped delta with distinct braided channels is clearly visible in the ortho-photographic map of the site and particularly after the application of a relief filter on the DEM of the area (Figs. 12-13). As these channels drain a lot of rainwater into the site, especially during heavy rainfall seasons, any ancient structure here has likely been completely eroded. Very probably the wadi existed long before any historical settlement and construction activities at J. Barkal, repeatedly eroding substantial amounts of sediment, while in other periods again accumulating alluvium. The resulting terrain is furrowed but relatively flat and even. 13 Wolf, P., ‘Recent Fieldwork at Musawwarat es-Sufra’, Sudan & Nubia 1 (1997), 20-29. Wolf, P. ibid and Scheibner, T., Wasserbauliche Infrastruktur und Wassermanagement in Musawwarat es-Sufra in kuschitischer Zeit, PhD, Vienna, 2017. 15 Well comparable to temple M282 (=KC102) in Meroe City; see Shinnie, P. and J. Anderson, The Capital of Kush, vol. 2, Meroë Excavations 1973-1984, in: Wenig, s. (ed.), Meroitica 20, Berlin 2004, pp. 44-56. 16 Both overlap also area JBC, surveyed in 2022, for several meters. 14 Report 23005 Page 20/30 Eastern Atlas GmbH & Co. KG Jebel Barkal Fig. 12 Intrusive wadi khor in the northwestern part of area JBF/JBH (ortho-photographic image) Fig. 13 Intrusive wadi khor in the northwestern part of area JBF/JBH (DEM with relief filter) Report 23005 Page 21/30 Eastern Atlas GmbH & Co. KG Jebel Barkal The central part of JBF/JBH features some of Reisner’s excavation dumps (b). They have been partially removed and sieved in recent years. A 9x12m large heap with potsherds (c) collected from these spoil dumps results from these recent activities. The surrounding terrain is characterized by silted-in shallow pits (k), indicating a clear anthropogenic influence (see below chapter 3.4.2.2). The southeastern third of area JBF/JBH returns to an even and flat terrain, interrupted only by some isolated shallow water gullies. It is one of the flattest areas within the southern part of the site at J. Barkal (Fig. 14). Fig. 14 Area JBF/JBH during the survey seen from south No excavations have yet been carried out in this area. However, a part of the southeastern half of JBF/JBH has been covered by magnetic gradiometry carried out by G. Tucker in 2016-2019. The magnetic gradiometry shows lots of densely packed anomalies in the area disturbed by the silted-in pits (k), but also in parts of the flat and level ground further southeast. Even before the GPR survey in 2023, this was a clear indication that anthropogenic subsurface remains were present here (Fig. 15). 3.4.2.2 Archaeological interpretation The most clearly recognizable sub-ground feature in the northwestern part of the area is a narrow trench (d) crossing it in north-south direction towards temple B300. According to S. Elamin, the trench was dug some years ago for a sub-ground electricity line to the temple. Report 23005 Page 22/30 Eastern Atlas GmbH & Co. KG Jebel Barkal Fig. 15 Results of former magnetic gradiometry surveys in the area of JBF/JBH and JBJ Directly in front of temples B200/B300, large but very weak reflections (f) can be identified. With some imagination, these could be interpreted as the foundations of a temple courtyard and a pylon with some minor structures in front of it. They roughly align with the orientation of B300-sub, the predecessor of temple B300 dating to the time of Ramses II. The dimensions of the structure that resembles a pylon of approximately 16x5 meters, match those of the other small New Kingdom temples at the site. However, it must be emphasized that this interpretation is highly speculative due to the weakness of the reflections. Just a few meters to the east, there are several large-scale reflections (g), whose orientation deviates by nearly 45° from that of the presumed temple pylon (f). This orientation matches that of a larger, seemingly square but completely destroyed and removed structure with an approximate side length of 35m. Its flat ruin mound is clearly discernible in both the ortho-photo base map and the 2006 magnetic gradiometry (Fig. 15). Based on its shape, it was likely a palace. It could be building B1150 or an as yet unidentified structure. While its orientation is relatively unusual compared to most of the structures at J. Barkal, the same orientation is also found in temple B200 and some reflections (h), located a few meters in front of this temple. These too are only very weak reflections, but they are visible at all depth levels. Reisner discovered and excavated a circular well with a diameter of 4m that had been cut partly into the sandstone bedrock and partly into the gravel in front of temple B200.17 It was located c. 10m to the "W" (of the B200 pylon) and "S" of the axis (of temple B200). Exactly at this location (i) (cf. Fig. 15 and Maps 23005-3.1. and 3.2), the GPR exhibits darker reflections, indicating higher amplitudes and thus a relatively moist ground environment. 17 “… about 10 meters to “W” and “S” of the axis is the circular well already mentioned cut partly in sandstone & partly in gravel.” (G.A. Reisner, Barkal-Kurru Diary 1918-19, 19-26 Jan. 1919 [pp 666-673], 666). Report 23005 Page 23/30 Eastern Atlas GmbH & Co. KG Jebel Barkal The area directly southeast of these structures, i.e. the central part of the west-east trending khor, is obviously devoid of any man-made structures, as they have probably been completely eroded. On the southeastern side of the khor, weak reflections (j) could indicate former brick structures. In this area, there is a scatter of dark reflections that could be attributed to scattered stone blocks or pits in the subsurface. A remarkable discovery was the detection of a previously unknown cemetery (k-l) in the central part of JBF/JBH (and partially in area JBJ). Approximately 220 grave pits clearly stand out from the surrounding ground due to higher GPR amplitudes. About 170 pits are arranged along seven east-west oriented rows (k), and approximately 50 pits form a separate cemetery area to the north (l), where the graves are laid out less orderly. The irregularly oval pits, 0.8-2.5 meters in length, are almost exclusively oriented north-south. Even without GPR, these pits are recognizable by their shallow depressions on the ground surface. No remains of grave superstructures have been preserved. Without excavation, dating and interpreting the cemetery is difficult. It could be the easternmost part of the Islamic necropolis located southwest of the site, as its graves have the same orientation. However, it is also possible that these are the pre-Islamic beginnings of the large cemetery. In this case, it could be a Medieval box grave cemetery or a Meroitic one. In both cases, the burial customs during these cultural periods would allow for a similar cemetery structure. Only a Napatan cemetery can probably be excluded, as the graves appear to cut into earlier (Napatan or Meroitic) structural remains. In the southeastern third of area JBF/JBH and in area JBI, where the erosive forces of the mentioned wadi no longer have any effect and where the present ground surface is so even that one would least expect to find archaeological remains here (cf. Fig. 14), the GPR detected underground structures with an extremely high density. In this maze of different shapes and orientations, it is hardly possible to differentiate clear structures from one another: Southeast of the above-mentioned cemetery, there are smaller-scale building structures (m), which correspond in their dimensions, wall thicknesses, and orientations with palace B1200 and the adjoining villas BP100 and BP200. Directly adjacent, a series of circular structures (n) seems to overlay earlier structural remains. The diameters of these round structures, ranging from 3.5-7m, suggest they are more likely the remains of round huts than pottery kilns or similar.18 However, they could also be remnants of round silos, such as those still built in the Nuba Mountains today. Adjacent to them, a very large rectangular building can be faintly discerned. Its exterior dimensions of appr. 40x50m and very thick walls suggest a medium-sized palace rather than a villa. If this interpretation is correct, its orientation would not conform with the other buildings in this area. This is, however, not unusual compared to other urban settlements like Meroe, where the street system was not strictly orthogonal as well (cf. Fig. 5). While the magnetic gradiometry carried out by G. Tucker outside the site fence shows significantly fewer structures, the high concentration and density of the structural features detected by GPR in this area suggest that the urban settlement extended beyond the current boundary of the site. Remains of enclosure walls were not detected at any location. 18 For comparable round structures associated with round huts see Wolf, P. and U. Nowotnick, ‘The Second Season of the SARS Anglo-German Expedition to the Fourth Cataract’, Sudan & Nubia 9 (2005), 23-31, esp. figs 2-3 and colour plate XVI, and Wolf, P. and U. Nowotnick, ‘The Third Season of the SARS Anglo-German Expedition to the Fourth Cataract of the Nile’, Sudan & Nubia 10 (2006), 20-32, esp. pls 6-8 and colour plate XVIII. Report 23005 Page 24/30 Eastern Atlas GmbH & Co. KG Jebel Barkal 3.4.3 Survey area JBJ (Map 23005-4) 3.4.3.1 General description Area JBJ forms a square of c. 53x55m with a little extension in the south (2531 m2), located immediately northeast of palace B1200. The area is mostly devoid of vegetation (with only a few grass bushes in the south-eastern part) but affected by many east-west running water gullies. It is generally horizontal but undulating due to underground features and test trenches dug by Reisner's workmen. There are also larger excavation dumps that had to be bypassed by the GPR. Its northeastern half is scattered by remarkably many column drums. Larger parts of the area had been surveyed with magnetic gradiometry in 2000 and again in 2006 (cf. Fig. 15). In addition, several test trenches were excavated by P. Wolf and U. Nowotnick in the central part of the area in 2002 to search for the remains of a coronation hall or temple, which T. Kendall had assumed to be here on the basis of the results of the gradiometry survey 2000. The trenches did not confirm the hypothesis, as only scattered stone debris was found, but no structural building remains. In 2006 a trench of 10x20m at the northwestern side of B1200 (located roughly in the northeastern quarter of area JBJ), which was not excavated by G. Reisner, was excavated by P. Wolf and U. Nowotnick in order to find the original pedestals of the so-called Prudhoe lions, which were originally located here according to Lord Algernon Percy Prudhoe's information, who visited J. Barkal in 1829 during his travels through Egypt and Sudan.19 In our excavation, we found the eastern corner of a larger building (a) with 1.5m thick outer walls made of unfired bricks. However, we were unable to identify the original location of the lions (Fig. 16). Fig. 16 Field drawing of the trenches excavated 2006 at the northwestern side of B1200 (© P. Wolf) 19 See Ruffle, J., ‘Lord Prudhoe and his Lions’, Sudan & Nubia 2 (1998), 82-87. Prudhoe had removed the lions during his visit to the site in 1829 and had given them 1835 to the British Museum London (Inv Nos EM 1 and EM2). Report 23005 Page 25/30 Eastern Atlas GmbH & Co. KG Jebel Barkal 3.4.3.2 Archaeological interpretation Despite the fact that area JBJ is disturbed by fluvial erosion, spoil heaps, and several test trenches from Reisner's excavations (c), the hitherto known northwestern walls of palace B1200 (d) were clearly detected. Even clearer GPR detected parts of a larger structure the eastern corner (a) of which was documented by the excavations in 2006 (cf. Fig. 16). Based on the GPR, we can now reconstruct a larger portion of it. It consists of at least two rooms, each measuring c. 10x8m. A nearly 5m wide doorway (e) seems to be in the southwest. However, such a width is extremely unusual for doorways in Kushite architecture. It is therefore likely that the structure extended further southwest (f), probably being part of the interiors of palace B1200. These core structures may have had a square shape (c. 60x60m), extending in the northeast approx. until (b). On the left and right of the supposed passageway (e), the GPR shows larger, nearly square reflections measuring c. 4x4m (g). It is conceivable that these mark the location of pedestals where the Prudhoe Lions were originally placed. If this is the case, the excavations in 2006 narrowly missed the original places of the lions. It is certainly due to the fluvial erosion by the wadi mentioned above that all structural remains further west in the JBJ area appear to have been destroyed. The many column drums and larger sandstone block fragments scattered over the area indicate that this structure, preliminarily designated B1150 by T. Kendall, may have been a larger building with substantial parts built in stone. It is therefore likely that area JBJ was the site of a coronation temple, as suggested by T. Kendall in 2006, a similar sacral monument, the eastern part of which may have been detected by the GPR and partially excavated in 2006, or that palace B1200 extended farther west having a square shape with dimensions of c. 60x60m. The magnetic gradiometry had identified in 2006 the square anomaly (h) of a presumably larger, but apparently completely destroyed building northwest of area JBJ (see chapter 3.4.2.2). This structure was not recognized by the GPR. 3.4.4 Survey areas JBK/JBL (Map 23005-5) 3.4.4.1 General description JBK and JBL, located in the area of the processional avenue in front of temple B500, comprise both together an L-shaped area. JBK is a northeast-southwest oriented rectangle of c. 40x63m (2496 m2), while JBL is an elongated, northwest-southeast oriented rectangle of 14x66m (950 m2) perpendicular to JBK. They cover roughly the southeastern half of the depression between temple B500 and the modern main track with its deepest point next to the track (JBL is located at the northeastern slope of it). JBK adjoins area A5/A4 of the GPR survey in 2020. Area JBK is crossed by many car tracks. Only a few larger grass bushes grew here and there is not much underground moisture in the two areas. JBK overlays temple B570 and JBL a part of temple B561. Both structures had been discovered by the magnetic gradiometry survey in 2006. Temple B561 and its associated kiosk B560 were excavated in 2014-2015 (see above chapter 1.1). 3.4.4.2 Archaeological interpretation Naturally, magnetic gradiometry can capture underground structures relatively roughly only, which is also true for the partially sandstone-built temples B570 (a), B561 (b), and B560 (c). As expected, the GPR resolved individual building features much more precise, at least in the case of temple B570. After more detailed examination of the data, it is even possible to interpret individual walls and scattered Report 23005 Page 26/30 Eastern Atlas GmbH & Co. KG Jebel Barkal building remains, such as (probably) several column drums southwest of the temple (d).20 However, it must be emphasized again that the interpretation is hypothetical, especially regarding the inner structure of the building and the construction materials used. B570 is likely a processional shrine, oriented perpendicular to the processional avenue, with external dimensions of approximately 35x23m, whose front is designed as a 28m wide pylon of 5.5m thickness. These features suggest that the temple dates to the Meroitic period. It likely had a transverse hypostyle hall of about 20x7m, a 7m broad vestibule, and a tripartite or a central sanctuary with ambulatory in the rear. Temples comparable in size, proportions and location along a processional way are e.g. KC 102 and M720 excavated by P. Shinnie in Meroe (Fig. 17)21 and Tempel 200 in Naga22. Fig. 17 Temples KC 100 and M720 in Meroe City (© P. Wolf) The situation is remarkably different for temple B561, whose hypostyle hall and vestibule lie in the northeastern part of area JBL (b): although this temple was fully excavated in 2014-2015 (cf. chapter 1) and is therefore covered by loose backfill, the GPR could hardly detect anything here. A few meters northeast of this temple, however, some structures are faintly visible (e-f), which, based on their orientation and dimensions – with due caution – could be interpreted as parts of further processional temples, which would be expected in these locations. The processional way proper, which must have been paved with stone slabs (referring to even much smaller sites like Hamadab), is also not clearly recognizable in the GPR, unless one interprets two lines running about 6-7m apart, but only very poorly visible, as its boundaries (g). Slight anomalies in the 20 They are relatively well recognizable in area A5 of the GPR survey in 2020. Cf. Shinnie, P. and J. Anderson, The Capital of Kush, vol. 2, Meroë Excavations 1973-1984, in: Wenig, s. (ed.), Meroitica 20, Berlin 2004, pp. 10-36. See also Wolf, P., Temples in the Meroitic South, Acta Nubica, Rome 2006, 236-262 for more information and literature. 22 Kuckerz, J. Naga - Temple 200 - The Wall Decoration, Archaeology in the Sudan ArS 2, Münster 2021. 21 Report 23005 Page 27/30 Eastern Atlas GmbH & Co. KG Jebel Barkal corresponding area are also visible in the magnetic gradiometry of 2006. The smaller and apparently thinner-walled, irregularly distributed structures (h) in the area southeast of temples B560/B561 and opposite temple B570 might possibly represent (later) squatter occupation. More detailed information about this significant sacred forecourt and the processional way of the largest Kushite temple must, as elsewhere, be provided by excavations. However, as the GPR and earlier magnetic gradiometry have shown, such excavations would certainly be archaeologically rewarding here. 3.5 Recommendations for future GPR prospection The results outlined in the previous chapters, along with those from the 2020 and 2022 seasons, underscore the importance of GPR in archaeological investigations, that offers a non-invasive method to uncover significant structural features of ancient settlements. The geophysical properties of the terrain at J. Barkal have proven to be particularly well-suited for GPR as a prospection method, while GPR and magnetic gradiometry complement each other effectively. Often, GPR detects structures in areas where magnetic gradiometry cannot, and vice versa. In areas where both methods were employed, GPR provided usually more precise and clearer results, especially in terms of the resolution of the surveyed structures. GPR is thus generally better suited than magnetic gradiometry. The high resolution achieved by GPR is particularly useful for creating a detailed map of the town, which can be drawn even without further excavations (a method successfully used at Hamadab). Moreover, the measurements in areas A4/5 in 2020 and in area JBG in 2023 illustrate that GPR devices with lower frequency antennas can penetrate the 2-3m thick spoil heaps from Reisner’s excavations with satisfactory prospection results. In the same time, the survey in areas JBA and JBB in 2022 showed that even in spots previously excavated by Reisner, GPR can still identify additional features and structures that lie stratigraphically below the excavated areas or represent ground features that early 20th-century archaeologists might have overlooked (e.g. planting pits). While magnetic gradiometry is more effective for large areas, given the ease of deploying devices with up to 10 sensors, GPR is more precise, though slower with current technology. This can be mitigated, however, by using multi-sensor devices. Since GPR is non-invasive and does not physically alter the archaeological site, it can also be used in areas outside the designated archaeological zones. For further investigation of the urban structure of Napata, it would therefore be desirable to expand GPR prospection to additional areas. Figure 18 shows a proposal for future prospection areas, with colors indicating different priorities. Initially, it would be reasonable to examine smaller areas, such as those covered by Reisner’s spoil heaps (A1 and A2) and extend the survey southwards (A3) to determine the extent of urban structures. Subsequently, examining the main track enclosing the site in the southeast (A4) to connect with the excavations and other areas on the Eastern Mound would be beneficial. If these surveys prove successful, the next priority would be to survey larger areas on the east side of Jebel Barkal (B1). Naturally, extending further north to the level of the Karima Museum (B2) would also make sense. As shown in areas JBF/JBH, flat terrain without specific archaeological surface features does not mean there are no man-made subsurface structures. On the contrary, the erodibility of materials like unfired bricks at J. Barkal has led to ancient buildings being entirely leveled by wind and water erosion. Ruin mounds survived only where stone blocks were used, but even these were often reused by the local population. Thus, the large areas B1 and B2 offer great potential for discovering more urban settlement structures of ancient Napata. Excavations by the Italian Mission at various Report 23005 Page 28/30 Eastern Atlas GmbH & Co. KG Jebel Barkal spots confirm that many remains likely lie beneath the surface. This is also corroborated by our own surveys, indicating that concentrations of mudbrick buildings are probably found where the ground is covered with whitish calcrete particles and small river pebbles. These aggregates were mainly used in Napatan palace buildings like B1200, whereas they were less used in later Meroitic constructions, as shown by excavations at the Eastern Mount. In a next step, it would be prudent to prospect larger open areas of Karima (C1) to determine how far the settlement extended northeast. Concurrently, it would be beneficial to use GPR in areas of the Eastern Mound that have not been surveyed by magnetic gradiometry (C2). It would also be worthwhile to re-survey areas previously examined with magnetic gradiometry using GPR. This would likely allow for the creation of a detailed city map without the need for complete excavation. In terms of area, the proposal covers just slightly more than what has been prospected so far. If we could achieve a detailed city map for this entire area by investing less time, effort, and funds than have been used for prospection until now, we should consider comparing it with the time, effort, and funds needed to excavate this area completely. This comparison would likely highlight the efficiency and cost-effectiveness of using GPR for extensive archaeological investigations at Jebel Barkal. Fig. 18 Proposal for future geophysical prospection Report 23005 Page 29/30 Eastern Atlas GmbH & Co. KG 4. Jebel Barkal Summary Napata's significance stems from its monumental temples and palatial structures, which suggest a substantial role as a religious and administrative center in the Kushite kingdom. Excavations by G. Reisner (1916-1920) and subsequent missions revealed much about the site’s sacral and political importance but provided limited evidence of a substantial urban settlement, leaving the exact location and extent of Napatan domestic structures unclear. While the GPR survey in 2022 had focused on areas around the Amun temple B500 and farther south between palaces B1200 and B100, the survey in 2023 investigated larger areas at the site in front of J. Barkal to the south of the Italian concession and apart from Reisner’s excavation dumps. It may be summarized as follows: Area JBF/JBH/JBI: Detection of large but weak reflections interpreted as foundations of a temple courtyard and a pylon, possibly aligned with B300-sub. Several structures indicating a larger, destroyed structure, likely a palace. Identification of approximately 220 grave pits, possibly part of the large Islamic cemetery or an earlier (pre-Islamic) cemetery. Detection of further villa-like resp. palatial structures adding data to the urban layout of early Napata. Area JBG: Larger structures next to the southwestern outer wall of courtyard B501 are possibly remnants of earlier sacred buildings. A group of round features likely to be interpreted as plantation pits of a temple garden, comparable to temple gardens in Musawwarat and Hamadab. Area JBJ: Northwestern outer walls of palace B1200 detected. Larger building structure identified, consisting of at least two rooms, with square features possibly marking the location of the Prudhoe Lions. The structure is potentially a coronation temple or palace B1200 extended farther north west, having a square shape with dimensions of c. 60x60m. Area JBK/JBL: While former magnetic gradiometry had captured only the rough structures of temples B570, B561, and B560, GPR resolved individual building structures much better, particularly in temple B570, suggesting it is a processional shrine dating to the Meroitic period. While not clearly recognizable in the GPR data, slight anomalies hint at the existence of a processional way. The GPR surveys of 2020, 2022 and 2023 demonstrated their value in detecting subsurface structures with high resolution and precision. Their results aligned with earlier magnetic gradiometry surveys, former excavations as well as test excavations carried out after the survey in 2023 to verify its results. By revealing detailed information about building structures and street networks, GPR has proven to be a crucial tool in locating and characterizing the town of Napata. The surveys, which had clearly identified urban structures by detecting several domestic, villa-like buildings, suggested a complex urban layout which is reminiscent of urban Kushite sites like Meroe, Hamadab, and Kawa. The internal structure and overall layout of the detected underground features indicate a densely populated area with complex building structures and streets, challenging previous notions of large open spaces around the monumental sacral and palatial buildings. This supports the presence of a well-organized urban settlement around J. Barkal, enhancing our understanding of its significance not only as a religious and administrative center but also as a densely populated urban settlement throughout the entire Kushite period. Report 23005 Page 30/30