The Iron Track of Jordan

J. Dik (Delft), G. van der Kooij (Leiden); also involved are: R. Hanssen, F. Dentz (Delft); H.A. Veldhuijzen (UCL), Z. Kafafi (Yarmouk University, Irbid, Jordan) through Leiden.

Fig.1 Southern Levant with Tell Hammeh

Geophysics in search of early iron production in Jordan

The Near East is considered to be the region where iron production was first developed around 1200 BC. The earliest example of a large-scale smelting site is Tell el-Hammeh at the junction of the Jordan and Zerqa Valleys (fig.1). The remains were discovered in 1996 and 2000 by joint excavations of Leiden University with Yarmouk University. More data had to be retrieved from the site, and earlier, more experimental forms of iron production in the region may have preceded the Hammeh one. Within CAAS two geophysical subprojects were designed: space-born remote sensing to search for additional iron production sites in the neighborhood, and a subsurface search by ground radar and magnetometry on Tell Hammeh, followed by excavations. Both subprojects are also pilot studies to specify the value of this kind of research for archaeology.

The satellite remote sensing subproject started January 2007, after satellite imagery was made available, as MSc study of Frank Dentz, supervised by Prof. Hanssen, and finalized January 2008 (fig. 2; cf. Geo-Info 5, 2008: 188-193).

Optical and radar satellite data were collected of an area of about 10x15 km, with height difference of 1000 m, including parts of the Jordan Valley, the Zerqa Valley (with Tell Hammeh and another site with slag) and the mountains east of the Jordan Valley, including the ore site of Mugharet Warde (cf. fig. 3).

Fig 2 Satellite imagery of the area
Fig 3. Tell Hammeh, left of centre, and at background hills with Mugharet Warde

The optical satellite images were multispectral (ASTER), hyperspectral (Hyperion) and high resolution multispectral (Quickbird). The radar data were also of three kinds. The images were studied and ground truth data were collected with a field spectrometer (April 2007) in order to establish geophysical ‘signatures’ of known smelting (slag) and ore sites. These were used to differentiate other ancient mining and smelting sites from the surrounding geology in the satellite imagery. It turned out that the mineral profiles were hardly recognizable in the satellite images. The known ore site was noted, but the signature differed little from the iron rich soil of the wider landscape. The slag’s signature was not recognized, partly due to the low surface density of slag remains. On the other hand geological formations and faults turned out to be identifiable from the images, as were tell-settlement sites and remains of irrigation canals, thus potentially contributing to archaeological studies of a little known region.

Fig. 4 Bakker explains handling of GPR tools

The non-invasive geophysical prospection and excavations of Tell el-Hammeh took place in Spring 2009, by Dr Marcel Bakker and Mr Mike Oude Munnink (geophysicists of Deltares, Utrecht), and archaeo-metallurgist Dr Xander Veldhuijzen. The prospection included Ground Penetrating Radar (GPR) and Magnetometry (MG). The 3D GPR records (fig. 4) of the internal structure of Tell Hammeh was expected to indicate architectural remains, actual iron production places, with slag and remains of furnaces, etc. However, due to the salinity and dense clay matrix of the soil, radar had a limited penetration, not more than some 60 cm, hardly sufficient to allow detection of metallurgical deposits. The MG provided better results, indicating locations with significant changes in the magnetic field. Based on the results of the two techniques, three excavation squares were opened, and it was found out that remains of stone walls correspond to the geophysical indications by MG. One GPR indication awaits further excavation.

Fig. 5 Veldhuijzen points at furnace debris, incl. a complete tuyere