ABSTRACT
ABSTRACT: Using geophysical methods for landscape-archaeology special requirements on speed, sensitivity and spatial resolution are needed. Due to a large variety of magnetization processes, magnetometry especially in a multi-sensor set-up meets these requirements in an optimal manner. The comparison of total fi eld caesium-magnetometry and fl uxgate gradiometry shows clearly, that caesium-magnetometry should be preferred because of its ultra high sensitivity. The detection of deeply buried archaeological structures with low magnetization contrast is only possible by this method. The key for this application was the development “from nanotesla to picotesla” in the mid 1990s: A caesium-magnetometer with picotesla sensitivity, heretofore used only in aeromagnetics, was put to use on the ground for archaeological prospection. Also during these years J. Faßbinder from the Munich lab discovered a new biogenetic magnetization process: So-called magnetic bacteria, which have built-in magnetite single domain crystals, are involved in the rotting of organic materials and the subsequent formation of soil. This means, that a wooden post (non-magnetic!) becomes a magnetic structure. The next step towards high speed (and high spatial resolution) was done by applying the time mode sampling. This laid to the use of the two sensors of the gradiometer (single track) as a duo-sensor magnetometer (double track, which doubles the speed of the sampling in the fi eld). By arranging the four sensors of two gradiometer systems horizontally one gets a quadro-sensor confi guration. A fi rst test of a quadro-sensor system was carried out in 1996 at Ostia Antica, the ancient harbour of Rome. Under smooth surface conditions the prospection of 1 hectare with 0.1/0.5 m spatial resolution can be done in 2 hours. Without these multi-sensor confi gurations it would have been quite impossible to measure huge sites like the Roman fort with vicus and necropolis near Ruffenhofen in Middle Franconia (c. one half square kilometre) in the short pauses available between the agricultural activities. But the largest area project was the prospecting at Qantir-Piramesse in Egypt from 1996 to 2004 reaching an over all area of nearly 200 hectares. Also the Celone Valley Project (2003-2006) in the Tavoliere (Apulia) with more than 100 hectares was only possible with the “magneto-scanner” with 4 sensors. Most of the projects also show, that geophysical prospection on the ground must be combined with other methods like aerial photography, laser scanning and high resolution satellite imagery for landscape archaeology. This was also demonstrated by the combination of magnetometry on the ground with Ikonos imagery at Uruk (Iraq) in 2001-02. The huge amount of data derived from high-speed magnetometry of large archaeological sites and for landscape archaeology every year arise new problems for data processing, visualization and archaeological interpretation. Still, there is a lot of work left in developing specialized software to transferring the information about the archaeological structures in the geophysical data to the human eye-brain system for interpretation, which cannot be done by computerized pattern recognition.
