Search

In archaeology, predictive models play a key role in understanding the interactions between humans and the palaeo-environment. They are also of great value for cultural heritage management and planning purposes. This is particularly true for Palaeolithic sites in the east Austrian loess landscape, which are often deeply embedded in sediment sequences. In this study, we analyse the geospatial behaviour of 23 Upper Palaeolithic sites in Lower Austria. Hereby, we apply a new approach, which combines the advantages of a classical deductive method with the capabilities of machine learning, implemented via the MaxEnt software. The result is a predictive model for an area of 7850 km², exploring the potential for the presence of Upper Palaeolithic sites. The model highlights several spatial dynamics of site probability in the study area. Possible sources of inaccuracies within the source data and the methodology are critically discussed.

In archaeology, predictive models play a key role in understanding the interactions between humans and the palaeo-environment. They are also of great value for cultural heritage management and planning purposes. This is particularly true for Palaeolithic sites in the east Austrian loess landscape, which are often deeply embedded in sediment sequences. In this study, we analyse the geospatial behaviour of 23 Upper Palaeolithic sites in Lower Austria. Hereby, we apply a new approach, which combines the advantages of a classical deductive method with the capabilities of machine learning, implemented via the MaxEnt software. The result is a predictive model for an area of 7850 km², exploring the potential for the presence of Upper Palaeolithic sites. The model highlights several spatial dynamics of site probability in the study area. Possible sources of inaccuracies within the source data and the methodology are critically discussed. [ABSTRACT FROM AUTHOR]