Volumes / Supplements / 8 / Cocca, Mandolesi

Analisi statistiche e geostatistiche con PyArchInit: prima sperimentazione

Enzo Cocca, Luca Mandolesi

«Archeologia e Calcolatori» 2016, Supplemento 8, 132-140

Abstract

The analysis of the spatial distribution of archaeological records combined with geostatistical techniques plays a primary role in the field of prehistoric archaeology because this methodological approach can be used to understand if a real or chance spatial correlation exists between archaeological records. This essay presents an approach to this issue which makes use of statistics, databases and GIS open source tools in a one-solution instrument based on QGIS desktop environment, in order to study an important archaeological site in Northern Italy (Fumane Cave, Verona). Since archaeology is an intrinsically spatial discipline, the team chose a spatial quantitative approach like the multivariate geostatistics (which is usually used for mining and environmental analysis), by managing the quantitative fieldwork information as regionalized variables. In this way, interesting spatial correlations were evaluated by imposing linear coregionalization models on experimental semivariograms and cross-semivariograms. Moreover, the team conducted this analysis by creating a tool that integrates several powerful R libraries such as Gstat for spatial distribution analysis of data, implemented in DBMS PostgreSQL with PostGIS extension.

Figures

View figures in Interactive Atlas of Digital Images

Preview

Subjects:

Statistics Survey and excavations

Download (PDF)

Publishers:

CNR - Istituto di Scienze del Patrimonio Culturale

Edizioni All'Insegna del Giglio