Christopher Sevara

Abstract

A shift in focus from the local to the regional is a hallmark of recent trends in the analysis and investigation of the present remains of the past. An integrated perspective, combining data from multiple sources in order to evaluate anthropogenic traces from a more holistic viewpoint, is another. Both of these trends are supported by advances in the collection and spatial resolution of non-invasive geospatial data, including the use of high resolution, large area remote sensing and geophysical prospection datasets. Additionally, new geospatial tools which take advantage of the high resolution of these datasets present opportunities to create bespoke applications which can aid in the rapid classification of the large amounts of data these approaches are producing. Parallel to these developments, the formalization of methodological and theoretical approaches in archaeology which move the evaluation of archaeological resources from discrete points on a map to a continuous coverage approach to representing change in the historical landscape has given the discipline new tools with which to explore the interrelated, ubiquitous and continuous nature of historical space. At the intersection of these developments lies an opportunity to bring these concepts together in order to create new ways to manage, analyze, classify and interpret the present remnants of past land use.

To that end, this project will focus on new, rapid approaches to creation of a continuous coverage geospatial dataset which characterizes the present remains of historic land use through the semi-automated detection and classification of remnant historic linear and areal landscape elements such as old roads and field boundaries. Using historic maps, aerial photographs, and other archival materials as structuring elements to determine correlations between relict boundaries of landscape components derived from integrated datasets, a relative time-depth and interpretation of the present remains of past land use will be created and will serve as a baseline for an analysis to determine the likely spatial and temporal relationships between different elements of the landscape above, at and below the modern ground surface. As an initial component of this project, new software tools which take advantage of recent developments in image analysis, terrain modeling and parallel computing will be developed to aid in the rapid categorization and extraction of these data. A main component of the theoretical framework of this study will be the analysis of the Historic Landscape Characterization (HLC) approach developed in the United Kingdom and the investigation of its applicability in the areas specific to this project.

The research design for this project consists of four key, interrelated goals:

• The creation of a Historic Landscape Toolkit, which will use geomorphometric, image based modeling and OBIA approaches to  aid in the rapid incorporation of historic data into project workflows through the use of semi-automated techniques for the positioning, classification and digitalization of historic landform elements.

• The development of a GIS-based multiresolution model through the integration of large scale geophysical and remote sensing datasets for the specific purpose of the characterization of remnant historic land use.

• The development of an integrated historic land use dataset, which uses HLC methods as a base for the creation of a spatiotemporal landscape model for the depiction of present remains of historic land use and estimation of historic landscape degradation.

• The investigation and application of geomorphometric landform classification methods as a tool for semi-automated remnant landscape classification and age-depth determination from integrated datasets.