Javier Fernández operating one of the drones used to map Sierra del Teleno. Photo credits: Javier Fernández Lozano.
High-res images enable scientists to build three-dimensional models of the site. Photo credits: Javier Fernández Lozano.
La resolución de sus imágenes aéreas es tan grande que se puede ver una moneda de dos euros en el suelo. Photo credits: Javier Fernández Lozano
La calidad de las fotografías permite la reconstrucción de modelos tridimensionales del terreno. Photo credits: Javier Fernández Lozano
Spanish researchers find vast Roman gold mines in Spain using geo-archaeological techniques
Researchers from the University of Cantabria, the University of Salamanca, the University of Castile-La Mancha and Complutense University of Madrid have joined forces in a geo-archaeological project to locate and map the largest gold mining complex in the Roman Empire, in Sierra del Teleno, Montes de León.
The area boasts a rich historical heritage which cannot be easily accessed, hidden as it is under thick vegetation and crops. Researchers had to use sophisticated detection methods and drones for digital reconstruction.
Javier Fernández Lozano (Madrid, 1982), PhD in Geology from Utrecht University (Netherlands), studied topography and photogrammetry at Madrid Technical University. He is also a drone pilot and professor at the Department of Sciences of the Earth and Condensed Matter Physics at the University of Cantabria. Fernández Lozano coordinated a study that unveiled the secrets of the Roman gold mining system in León using the most sophisticated techniques available.
- You have mapped the largest Roman mining complex across Europe in Montes de León. Could you describe your project for us?
- We used detection technology mounted on drones to explore 30,000 hectares. We took high-resolution photographs that can show a 2-euro coin from the sky.
What we did with drones had never been done before. We used drones to make a cartographic model of the vast ancient gold mines in Spain, which were not unknown but had never been approached as we did. We introduced new methodology, and our data will be available for other scientists to analyse the Roman gold mines in Spain in depth.
- The area cannot be easily accessed, so your study can help discover new characteristics of the Roman gold mines in Spain.
- That is right. It is rugged terrain, and getting close to the ground is really difficult, so taking photos is not easy. You can take photos from the sky but they are low-resolution photos. Digital models do not deliver high-quality images to work on.
Now, with drones, you can take high-resolution pictures, edit them and get huge amounts of data. With the edited images, we located the remains of canals, water reservoirs and mining areas. Then we processed the data to construct a cartographic model that will enable us to analyse and describe archaeological remains.
- You also resorted to photogrammetry. How would you define it?
- Photogrammetry uses modern techniques, but it is an ancient discipline. It has its origins in the discovery of perspective by Leonardo da Vinci, back in the fifteenth century. In the nineteenth century, we were able to get sequences of aerial images. At the outbreak of World War II, a photogrammetric technique was developed that consists in reconstructing an area’s relief using aerial photos. At first, the technique was used for military purposes; then it extended to topographic maps.
We used a photogrammetric range imaging technique known as ‘Structure from motion’ for estimating three-dimensional structures from two-dimensional image sequences that may be coupled with local motion signals. The study areas were identified on the basis of pixels and software.
This technique provided us with point cloud data that can be used to make digital models. With three photos, you can go from a two-dimensional to a three-dimensional model.
- What are the advantages of this methodology?
- The main advantage is that both airborne laser (to erase the vegetation) and drones produce huge amounts of data for detailed scientific description. Thus, we can understand what the gold mines really looked like, in all their glory.
Our methodology offers a solution to a problem often faced by archaeologists: the lack of document records, either because they have not been able to dig or they have not found them, or because there are no literary or epigraphic sources to reconstruct the mining site.
- How long have you been working on this project?
- Four years. We have covered 30,000 hectares so far, and we would like to carry on. It can take weeks for the models to be developed.
Think about this: each image is 6 to 10MB. There are 300 or 400 images per model – in some models, up to 1200 images. Do the maths… The amount of information is humongous!
- What can you tell us about these ancient gold mines so far?
- The mines have been exploited non-stop since the Romans. There are many experts who are familiar with the site. We have extended coverage for the record, offering a new ultra-high-res geometry. Of course, this has great value for the scientific community. But it is also important for the general public: our high-res pictures show the people their own archaeological heritage and, in so doing, engage locals in its protection and preservation.