Author: Jan Fišer
The history of the use of certain raw materials goes back thousands of years. This is also the case for the quartzite deposits at Písečný vrch near Most, Czech Republic. While in prehistoric times raw material was used to make stone tools, several thousand years later – in the 20th century – it found use as a raw material for the production of semiconductors. This layering of mining activities has left a mosaic of anthropogenic traces in the landscape – pits, quarries, spoil heaps, and trenches. However, many of these pits may originate from different periods, so our goal was to decipher this mining palimpsest and distinguish between prehistoric and modern interventions in the terrain.
MaxEnt modeling and variables.
For the analysis, we used a combination of field surveys and remote sensing data (historical aerial photographs, digital terrain models (DTM), and historical cadastral maps). Using data obtained from the field survey and historical aerial photographs, we established a reference layer consisting of pits that could be reliably attributed to the 20th century. This dataset was then used as a presence-only input for the MaxEnt (Maximum Entropy) predictive model. The result value of the modeling was the so-called Recency Index, which expresses the likelihood that a given pit was created during the 20th century. This index thus allows for a relative estimate of the age of individual pits within the entire set. Higher Recency Index values were achieved mainly by larger pits with exposed quartz in the walls, which corresponds to the method of modern mining – i.e., excavating and breaking out larger blocks.
How many pits are recent?
Using this approach, we were able to evaluate 656 pits and classify them according to their archaeological potential. A total of 331 pits were identified as likely to be older, i.e., with potential for further archaeological research. Our study highlights the importance of combining multiple data sources and predictive modeling to distinguish layers of overlapping human activities. Remote sensing data provided a detailed overview of the entire study area, but field survey was also an essential part of the process, allowing us to identify recent pits, for example, based on traces of modern tools which cannot be distinguished from remote sensing data. Although the data used has its limitations, this approach provides a suitable framework for the systematic and non-destructive differentiation between prehistoric and recent pits and opens up new possibilities for further archaeological and landscape-historical research.
