The research in Soba is funded by the National Science Centre, Poland (Grant Number UMO-2018/29/B/HS3/02533) and hosted by the consortium of the Polish Centre of Mediterranean. Archaeology of the University of Warsaw and the Institute of Archaeology and Ethnology of the Polish Academy of Sciences. The authors would like to sincerely thank to Jarosław Majewski (gpr.software) for enabling the use of Wave software. Our work in Sudan would not be possible without the permission and support of the National Corporation of Antiquities and Museums in Khartoum.
1The survey forms part of an interdisciplinary project investigating the spatial organisation of medieval Soba, the capital of Kingdom of Alwa (Drzewiecki & Ryndziewicz, 2019). Located on the right bank of the Blue Nile, approximately 20km south of the confluence of the White Nile and the Blue Nile in Sudan, the city was subject of archaeological research in the 20th century, which unearthed remains of monumental architecture dated from 6th to 14th century (Shinnie, 1961; Welsby & Daniels, 1991; Welsby, 1998; Abdel Rahman, 2000). Historical records indicate that Soba was the administrative heart of the kingdom, and together with the kingdoms of Makuria and Nobadia formed three Christian realms covering entire Nubia. Soba was thus an important religious centre with bishopric see and commercial hub.
2Today, its remains are located in the eastern suburbs of a much younger capital city – Khartoum, a rapidly developing African metropolis. It has been estimated that at its peak Soba covered the area of approximately 275 ha, yet only about 50 ha of the archaeological site remain available for survey. Historic satellite imagery (Fig. 1) shows that as late as in the 1970’s, the area was broadly undeveloped with patches of vegetation characteristic to the semi-arid conditions of the Sahel and an intermittent strip of arable fields abutting the Nile. However, the urbanization processes have intensified in recent decades leaving the site, densely built-up by modern houses with areas closer to the river turned into fields, pasture and orchards. The site has also been intersected by a tarmac road with heavy traffic throughout the day.
Figure 1. Multitemporal landscape changes analysis showing the development of modern infrastructure in Soba.
Data source: A – Corona satellite imagery, B, C, D – Google Earth. Coordinate system: WGS84 / UTM zone 36N.
3Except for the remains of walls unearthed during previous excavation campaigns, the up- standing remains of stone columns and several mounds covered with red brick rubble, no remnants of medieval buildings are visible in the archaeological landscape of Soba.
4While it is difficult to scale up conventional archaeological exploration methods to cover large areas, the current situation in Soba requires mapping of threats and damage to the site, which could be then used to develop a proposal of policy for preserving the archaeological heritage. For this reason, a set of non-invasive procedures was used to investigate the area. This aimed to obtain large-scale high-resolution spatial data, which was later used to characterize the attributes of archaeological remains not visible on the surface, and detect and map features such as squares, streets, districts or the layout of individual buildings.
5The non-invasive research was basing primarily on the magnetic method. For the measurements, we used Bartington Grad 601-2 fluxgate gradiometer, with 0.1 nT resolution, within 0.5 × 0.25 m sampling grid. The data was processed using Geoplot 4.0 software. While it has been planned to cover 50 ha of the site that are still available to survey, to date, work has been completed on approximately 20.5 ha. In addition, UAV were flown to collect imagery used for detailed photogrammetric documentation of the surface (generating orthomosaics) and to produce digital surface model of the site. Selected areas were also investigated using Ground Penetrating Radar – MALA GX system equipped with a 450 HDR antenna. Measurements were carried out along parallel traverses set 0.5 m apart, covering a total area of approximately 4.4 ha. Raw data obtained from these measurements were processed using Wave software. Positioning data, collected using a RTK GNSS unit (Topcon Hiper SR) allowed for integrating survey outputs in GIS using QGIS software, and combine it with other data sets, such as archival cartographic documentation, plans and satellite imagery. This facilitated data management at subsequent stages of the survey and the ongoing evaluation of the outputs.
6The outputs of geophysical surveys should be considered as a combination of geological, environmental, and anthropogenic factors. While the use of two geophysical methods allowed for obtaining comprehensive information, the existing conditions affected each method. The magnetic survey (Fig. 2), which is often used in the Nile Valley (e.g. Herbich & Ryndziewicz, 2019), allowed to locate the remains of mud-brick buildings. Some anomalies can be interpreted as the remains of red brick features, furnaces and ash dumps. Voids identified between the densely built-up areas, where the minimal magnetic response revealed almost no features, may offer a view on the spatial organization of Soba or result from the limitations of the methods used (for example in recording traces of timber architecture). In some areas, the impact of ploughing on the architectural remains is clearly visible. Red brick scattered on the surface of several mounds caused high-intensity anomalies, completely occluding archaeological interpretation of any structures below.
Figure 2. Results of the magnetic survey undertaken in 2019/2020 season in Soba. In enlarged areas A, B and D, the results are interpreted as the remains of dense mud brick architecture in a compact urban layout. The results for area C are interpreted as a building complex erected from mud brick and red brick.
Coordinate system: WGS84 / UTM zone 36N.
7While GPR survey (Fig. 3) was carried on a limited scale, aimed to obtain detailed information of sub-soil architectural remains, the most significant limitation turned out to be very strong attenuation of the electromagnetic wave, limiting the depth range of the prospection to about 1-1.2 m with the relative dielectric permittivity approximately set at 5.75. The stratigraphy of features can be understood from the collected data, in some cases however, it turned out to be difficult in the existing survey conditions.
Figure 3. Extract from the GPR survey showing the amplitude time slice between 9ns and 10ns (0.54 m to 0.60 m) of area shown on Fig. 2C superimposed over digital surface model.
Coordinate system: WGS84 / UTM zone 36N.
8Non-invasive research demonstrated its potential to improve understanding of the spatial organization of Soba by providing a large and comprehensive set of information. Geophysical data interpretation achieved a higher level of reliability when combined with other datasets and methods including limited small-scale excavations. The obtained data made it possible to specify areas requiring protection and management.