University of Southampton OCS (beta), CAA 2012

Font Size: 
Landscape reconstruction and archaeological prediction in the Lake Manyara area, Northern Tanzania
Michael Märker, Geraldine Quénéhervé, Felix Bachofer, Christine Hertler, Liane Giemsch

Last modified: 2011-12-17


In this study, we focus on archaeological sites within the Makuyuni River basin, draining into Lake Manyara, in Northern Tanzania. This region is known for Middle and Lower Pleistocene fossil finds and artefacts. Non-invasive field methods were used to characterize environmental variables for the prediction of physiographic conditions and archaeological sites. In particular, detailed information, especially on the distribution of soils and substrates, are measured along typical transects. We applied advanced geo-statistical methods to regionalize the given information. The methodology considers not only the characteristic of specific transects but implicitly also the related pedogenetic and morphogenetic processes including actual erosion, transport and deposition processes.

To analyze the spatial distribution of potential archaeological and paleontological find locations we applied a methodology based on statistical mechanics. This method is able to handle presence-only datasets such as those are. The spatial distributions of the find locations were collected in literature and by own field work over the last years. For the modelling, we utilized environmental information such as spatially continuous layers of soils and substrates, topography (30m SRTM DEM), vegetation information, as well as ASTER multispectral data as predictor variables. The results reveal the potential spatial distribution of the physiographic process dynamics and the related distribution of paleontological/archaeological sites.


prediction modelling; presence-only data; spatial distribution; predictor variables