Weights of Evidence Predictive Modelling in Archaeology
(2022) In Master Thesis in Geographical Information Science GISM01 20221Dept of Physical Geography and Ecosystem Science
- Abstract
- Predictive archaeological modelling is a complex analytical process that requires the understanding of how complex environmental and anthropogenic spatial phenomena relate to the selection of archaeological site location and knowledge of the requirements and potential biases inherent in the method of modelling and the raw data itself. This paper aims to demonstrate the utility of the weights of evidence method for predictive modelling via the ArcSDM toolkit for ArcGIS of Bronze Age settlement patterning in two river valleys in Cyprus.
The weights of evidence method is a probability-based procedure for determining archaeological potential using the spatial distribution of indicative features with respect to known archaeological site... (More) - Predictive archaeological modelling is a complex analytical process that requires the understanding of how complex environmental and anthropogenic spatial phenomena relate to the selection of archaeological site location and knowledge of the requirements and potential biases inherent in the method of modelling and the raw data itself. This paper aims to demonstrate the utility of the weights of evidence method for predictive modelling via the ArcSDM toolkit for ArcGIS of Bronze Age settlement patterning in two river valleys in Cyprus.
The weights of evidence method is a probability-based procedure for determining archaeological potential using the spatial distribution of indicative features with respect to known archaeological site locations. Weights (W+, W−) and contrast (C=(W+)−(W−)) calculations guide the data-driven procedure. The model uses site location data taken from two methodologically distinct ground surveys and six “evidential layers” reflecting natural and anthropogenic spatial phenomena (hydrogeology, rivers, vegetation, landuse, soil and slope) to produce predictor maps for the study region.
The purpose of these maps is twofold. They aid academic study by depicting changes and trends in settlement patterning that occur over the course of the Bronze Age and can be analysed against current theoretical debates on socioeconomics and they inform land development policy in the commercial sector by identifying archaeological sensitive areas that should receive due care.
The analysis of these relationships using GIS and weights of evidence modelling yielded 51km2 of archaeological favorable landscape, 26 specific locations within 8 sub-regions of particular importance and guidance that led to a successful ground survey in a previously unsurveyed region of the study are that led to finds spanning the Chalcolithic to modern era. (Less) - Popular Abstract
- Locating archaeological sites is a time-consuming, expensive, and difficult process due to the high expenditure of resources involved in field work. Desk-based GIS predictive archaeological modeling has become a cost effective and efficient tool for identifying areas of archaeological significance. This is accomplished by comparing the spatial attributes of the landscape that co-occur at known sites and using this information as a recipe for areas where, yet unknown archaeological significance might reside. Some methods for archaeological predictive modeling are better than others. This thesis aims to demonstrate that the Weights of Evidence method is particularly adept at identifying and factoring into predictions the statistical... (More)
- Locating archaeological sites is a time-consuming, expensive, and difficult process due to the high expenditure of resources involved in field work. Desk-based GIS predictive archaeological modeling has become a cost effective and efficient tool for identifying areas of archaeological significance. This is accomplished by comparing the spatial attributes of the landscape that co-occur at known sites and using this information as a recipe for areas where, yet unknown archaeological significance might reside. Some methods for archaeological predictive modeling are better than others. This thesis aims to demonstrate that the Weights of Evidence method is particularly adept at identifying and factoring into predictions the statistical uncertainty and various sources of bias inherent in predictive modeling.
Specifically, this thesis demonstrates the utility of the weights of evidence method for predictive modelling of Bronze Age settlement patterning in two river valleys in Cyprus using the ArcSDM toolkit for ArcGIS. The model uses known site location data taken from two archaeological surveys and six landscape attributes (hydrogeology, rivers, vegetation, landuse, soil and slope) to produce archaeological significance predictor maps for the study region.
The final predictor map identified 51km2 of archaeologically significant landscape. One area of approximately 10 km2 that was ground surveyed to verify the predictor maps was found contain significant archaeological material that spanned almost 5000 years of occupation. Moving forward, the predictor maps will serve two important roles. They can aid academic study by showing a more complete view of the Bronze Age archaeological landscape which can be used to test current theories on settlement patterning and socioeconomics. These maps can also inform land development policy in the commercial sector by identifying archaeological sensitive areas that should receive due care. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9074317
- author
- Aspland, Luke LU
- supervisor
- organization
- course
- GISM01 20221
- year
- 2022
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Archaeology, Physical Geography and Ecosystem analysis, GIS, Weights of Evidence Predictive modelling, evidential layers, predictor map
- publication/series
- Master Thesis in Geographical Information Science
- report number
- 141
- language
- English
- id
- 9074317
- date added to LUP
- 2024-06-04 12:31:28
- date last changed
- 2024-06-04 13:30:21
@misc{9074317,
abstract = {{Predictive archaeological modelling is a complex analytical process that requires the understanding of how complex environmental and anthropogenic spatial phenomena relate to the selection of archaeological site location and knowledge of the requirements and potential biases inherent in the method of modelling and the raw data itself. This paper aims to demonstrate the utility of the weights of evidence method for predictive modelling via the ArcSDM toolkit for ArcGIS of Bronze Age settlement patterning in two river valleys in Cyprus.
The weights of evidence method is a probability-based procedure for determining archaeological potential using the spatial distribution of indicative features with respect to known archaeological site locations. Weights (W+, W−) and contrast (C=(W+)−(W−)) calculations guide the data-driven procedure. The model uses site location data taken from two methodologically distinct ground surveys and six “evidential layers” reflecting natural and anthropogenic spatial phenomena (hydrogeology, rivers, vegetation, landuse, soil and slope) to produce predictor maps for the study region.
The purpose of these maps is twofold. They aid academic study by depicting changes and trends in settlement patterning that occur over the course of the Bronze Age and can be analysed against current theoretical debates on socioeconomics and they inform land development policy in the commercial sector by identifying archaeological sensitive areas that should receive due care.
The analysis of these relationships using GIS and weights of evidence modelling yielded 51km2 of archaeological favorable landscape, 26 specific locations within 8 sub-regions of particular importance and guidance that led to a successful ground survey in a previously unsurveyed region of the study are that led to finds spanning the Chalcolithic to modern era.}},
author = {{Aspland, Luke}},
language = {{eng}},
note = {{Student Paper}},
series = {{Master Thesis in Geographical Information Science}},
title = {{Weights of Evidence Predictive Modelling in Archaeology}},
year = {{2022}},
}