EAA2020: Abstract
Abstract is part of session #241:
Title & Content
Title:
Spatiotemporal interpolation based on bulk radiocarbon data: Two case studies from archaeology and archaeogenetics
Content:
The amount of available radiocarbon dates from archaeological contexts is enormous and openly accessible data collections like EUROEVOL [1], AustArch [2] or KITEEastAfrica [3] (just to name a few) are a future gold mine for meta-studies. One interpretation of this data is especially useful for the reconstruction of long-term and large-scale processes in prehistory: Radiocarbon dates can be seen as independent "measurements" in conjunction with dependent variables. The dependent variable can include any kind of cultural or natural data relevant for a specific research question. Modelling it in relation to the independent radiocarbon sample position in space and time is, therefore, a promising meta-analysis method with a huge spectrum of applications. We present two independent case studies where we apply Gaussian Process Regression on contextualized radiocarbon samples to interpolate the distribution of Bronze Age burial rites and human genetic ancestry components.
Case study 1: European Bronze Age archaeology traditionally focuses on two major dimensions to categorize burials — although there is an immense variability of attendant phenomena: Flat graves versus burial mounds and cremation versus inhumation. The radiocarbon database RADON-B [4] contains dates on graves attributed to these major categories. We reconstruct their diachronic development in Central- and Northwestern Europe and distinguish cultural units with homogeneous history.
Case study 2: A recently compiled global archaeogenetic dataset [5] combines published genotype information for ancient and present-day individuals with context information about the position in space and time for each sample — the latter most often derived from radiocarbon dates. The reported 597,573 genome-wide ancestry-informative markers are sufficient to quantify the genetic relationships between all individuals using multivariate statistics such as Principal Component Analysis. We use these methods to derive basic ancestry components for European prehistory, interpolate their spatiotemporal distribution and explore patterns of large-scale change.
Case study 1: European Bronze Age archaeology traditionally focuses on two major dimensions to categorize burials — although there is an immense variability of attendant phenomena: Flat graves versus burial mounds and cremation versus inhumation. The radiocarbon database RADON-B [4] contains dates on graves attributed to these major categories. We reconstruct their diachronic development in Central- and Northwestern Europe and distinguish cultural units with homogeneous history.
Case study 2: A recently compiled global archaeogenetic dataset [5] combines published genotype information for ancient and present-day individuals with context information about the position in space and time for each sample — the latter most often derived from radiocarbon dates. The reported 597,573 genome-wide ancestry-informative markers are sufficient to quantify the genetic relationships between all individuals using multivariate statistics such as Principal Component Analysis. We use these methods to derive basic ancestry components for European prehistory, interpolate their spatiotemporal distribution and explore patterns of large-scale change.
Keywords:
Radiocarbon, Spatiotemporal modelling, Archaeogenetics, Bronze Age, Burial rites
Downloads:
authors
Main authors:
Clemens Marcus Schmid1
Co-author:
Stephan Schiffels1
Affiliations:
1 Department of Archaeogenetics, Max Planck Institute for the Science of Human History, Jena, Germany