Archaeological prospection of forested areas using full-waveform airborne laser scanning

https://doi.org/10.1016/j.jas.2007.06.013Get rights and content

Abstract

Airborne laser scanning (ALS) is a potential tool for recognising and measuring topographic earthwork features in wooded areas. To explore its potential for archaeological reconnaissance in a densely forested area, a test scan covering an Iron Age hillfort in the eastern part of Austria was carried out during the first phase of a research project.

ALS sensors can penetrate vegetation canopies allowing the underlying terrain elevation to be accurately modelled. The latest generation of airborne laser scanners was used in the project. This sensor digitally records the entire waveform of the received laser echoes. We argue that the digital terrain model (DTM) generated from entire waveform ALS data could be classified with greater confidence providing a more accurate DTM than with previous ALS devices. The processing algorithms used to create the interpretative DTM are discussed in detail.

Using the described procedures it was possible to remove most of the forest canopy and understorey (brushwood and low level vegetation) covering the archaeological features. The ALS DTM was compared with a detailed topographic mapping of the visible archaeological traces collected by a terrestrial survey. Significantly, very low earthwork features, which were not recognized by the trained surveyors in the field, were identified in the ALS-derived DTM. Therefore, in this study area ALS has been demonstrated as an important tool for systematic archaeological prospection in vegetated areas. There are, however, some restrictions, which are discussed in the paper.

Introduction

Despite the success of archaeological prospection in agriculturally dominated regions, the identification of sites within wooded areas remains problematic. At present there is no prospection method for the systematic discovery of buried sites in forests. Fortunately, micro and macro topographic earthwork features tend to be well preserved in forested areas, due to the stabilizing effect of vegetation on erosion processes and the lack of surficial disturbance through mechanical action (such as ploughing).

Large (macro topographic) earthworks can be recognised from the air (using optical aerial photography techniques) typically as shadow marks or snow marks (Wilson, 2000, p. 38), even through the tree canopy. They are also visible from the ground and can be mapped using traditional surveying techniques: however, a woodland environment can make survey difficult and features may be obscured by brushwood and scrub. Conversely, low (micro topographic) earthworks, especially when covered with dense vegetation, are practically invisible from the air and can be difficult to locate on the ground even by an experienced surveyor.

To be able to facilitate identification of low earthwork features, one would need a dense coverage of surface terrain points that are accurately located in the x, y and z dimension. These points are used to generate a Digital Terrain Model (DTM) that can be used for archaeological interpretation. Only a few years ago, the measurement of such a huge quantity of points would have been impossible for larger areas, but with the recent development of Airborne Laser scanning (ALS) there is now the means to produce dense, precise, and accurate terrain models (Ackermann, 1999, Kraus, 2004, pp. 449–470; Wehr and Lohr, 1999) even under forest canopy (Kraus and Pfeifer, 1998, Pfeifer et al., 1999). In this paper an approach employing ALS techniques is presented.

There are an increasing number of ALS-applications in archaeology, but investigations in forests are rare (Devereux et al., 2005, Harmon et al., 2006, Sittler, 2004, Sittler and Schellberg, 2006, Risbøl et al., 2006). All published examples used terrain models derived from data collected by conventional ALS. We argue that conventional ALS may not be able to resolve the difference between near ground vegetation and the underlying terrain which can reduce the quality of any resultant DTM (Pfeifer et al., 2004). This inhibits any subsequent archaeological identification of low earthwork features.

This paper will present the latest generation of full-waveform recording ALS systems. It is hypothesised that these sensors have a number of advantages, especially in vegetated areas, which can lead to a better identification of low earthwork features. After introducing the basic techniques of ALS processing, the paper will deal with the under-represented aspect of deriving an archaeologically relevant digital terrain model (DTM) from unfiltered ALS data. The value of full-waveform ALS data is illustrated with a case study of an Iron Age hillfort in the eastern part of Austria.

Section snippets

Principle of ALS

ALS, also referred to as LiDAR (Light Detection and Ranging), is an active remote sensing technique (Wehr and Lohr, 1999). The laser scanner is usually mounted below an aeroplane or helicopter, where it emits short infrared pulses into different directions across the flight path towards the earth's surface (typically 30,000–100,000 pulses per second). Each pulse will result in one or more echoes reflected from various objects along its path (vegetation, buildings, cars, ground surface etc.).

Case study – the Iron Age hillfort of Purbach

The Iron Age hillfort consists of linear and non-linear above ground earthwork features with varying preserved heights. It is covered by a forest with varying degrees of understorey, but contains cleared parts which are partly overgrown with dense bushes and partly covered with clearance piles. It represents a range of environments and therefore it is an ideal case-study to test the applications of full-waveform ALS data.

ALS and terrestrial surveying

The comparison provokes some general thoughts on ALS and terrestrial surveying (Doneus and Briese, 2006b). During terrestrial survey, interpretation is an implicit element of recording (i.e. an object is only recorded if the survey has deemed it to be significant). While interpreting, the archaeological surveyor is literally in touch with a site which is already known. This has many advantages; for example, it will not be difficult for the surveyor to distinguish a pile of wood from a barrow.

Conclusion

The paper demonstrated the potential of full-waveform airborne laser scanning for archaeological prospection of forested areas. An Iron Age hillfort with various ramparts and round barrows hidden in a forest, with varying structure of trees and bushes worked as a test site. The area was scanned during early spring, when the deciduous trees were still without leaves.

For archaeological interpretation, a high quality DTM has to be derived from the ALS data. This involves a reliable separation of

Acknowledgments

This research has been supported by the Austrian Science Fund (FWF) under project no. P18674-G02. Figs. 8b and 11 are printed with the kind permission of the Austrian Federal Office of Metrology and Surveying (Bundesamt für Eich- und Vermessungswesen). The authors also want to thank Thomas Melzer from the Christian Doppler Laboratory for Spatial Data from Laser Scanning and Remote Sensing, Vienna University of Technology, for the processing of the Gaussian decomposition of the full-waveform ALS

References (35)

  • M. Doneus et al.

    Digital terrain modelling for archaeological interpretation within forested areas using full-waveform laserscanning

  • M. Doneus et al.

    Full-waveform airborne laser scanning as a tool for archaeological reconnaissance

  • EuroSDR
  • J.M. Harmon et al.

    LIDAR for archaeological landscape analysis: a case study of two eighteenth-century Maryland plantation sites

    American Antiquity

    (2006)
  • N. Holden

    Digital airborne remote sensing: the principles of LIDAR and CASI

    AARGNews

    (2001)
  • N. Holden et al.

    High-resolution digital airborne mapping and archaeology

  • C. Hug et al.

    Litemapper-5600 – a waveform-digitizing LIDAR terrain and vegetation mapping system

  • Cited by (217)

    • Interpolation of airborne LiDAR data for archaeology

      2023, Journal of Archaeological Science: Reports
    • Earth observation in archaeology: A brief review

      2023, International Journal of Applied Earth Observation and Geoinformation
      Citation Excerpt :

      Over the past century, sensing-based EO has seen a large number of innovative archaeological applications and encountered some emerging and challenging issues that may not been covered in existing publications. Some reviews, articles, commentaries, perspectives, book chapters and technical notes (Davis, 2018; Doneus et al., 2008; Lasaponara and Masini, 2011; Luo et al., 2019; Masini et al., 2018; McCoy and Ladefoged, 2009; Menna et al., 2018; Opitz and Herrmann, 2018; Tapete, 2018) that pick up and identify the fundamental ideas and beneficial practices that demonstrate different sensing-based EO techniques applicable for archaeological purposes and contribute to fruitful achievements have been published. However, most of these concentrated on the aerial and satellite RS platforms; a comprehensive retrospect has not been conducted and an insightful prospect on progressions, challenges and future directions is insufficient or even lacking.

    View all citing articles on Scopus
    View full text