close

Вход

Забыли?

вход по аккаунту

?

Initial results using GPS navigation with the Foerster magnetometer system at the World Heritage site of Cyrene Libya.

код для вставкиСкачать
Archaeological Prospection
Archaeol. Prospect. 15, 151–156 (2008)
Published online in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/arp.330
Short Report
Initial Results using GPS Navigation
with the Foerster Magnetometer
System at theWorld Heritage Site of
Cyrene,Libya
CHRIS GAFFNEY1*,VINCE GAFFNEY2,
RICHARD CUTTLER2 AND RON YORSTON3
1
Archaeological Sciences, Division of Archaeological, Geographical and Environmental
Sciences, University of Bradford, Bradford,West Yorkshire BD7 1DP, UK
2
The Institute of Archaeology and Antiquity, Arts Building, University of Birmingham,
Edgbaston, Birmingham B15 2TT, UK
3
1ChurchTerrace, Lower Field Road, Reading RG1 6AS, UK
ABSTRACT
This Short Report summarizes some initial results using real time GPS to navigate and collect magnetometer data using Foerster sensors and a magnetic cart.The Foerster system is primarily aimed
at the detection of buried ordnance and, by comparison to some other magnetometer sensors, the
reported sensitivity is relatively low. However, the sensors require no alignment in the field, nor does
the system require a regular contiguous grid to be established across the survey area. The latter
meansthat data grids of different sizes and orientation can be measured and stitched togetheraspart
of data restoration prior to processing. The accurate positioning of the data means that ‘staggering’,
whichis often seenintime-based collection strategies, isnot apparent in the data; overalllessprocessing is required to produce a final image than is required for other fluxgate instruments. Data from a
surveyat Cyrene arereproduced toillustrate the strengthsofthe system.Copyright # 2008 JohnWiley
& Sons, Ltd.
Key words: fluxgate magnetometer; Foerster; real time GPS; de-striping; topography; Cyrene
Introduction
The majority of magnetometer surveys undertaken for archaeological purposes require highdensity data and, in order to establish accurate
positioning, the survey area is normally divided
* Correspondence to: C. Gaffney, Archaeological Sciences,
Division of Archaeological, Geographical and Environmental
Sciences, University of Bradford, Bradford, West Yorkshire
BD7 1DP, UK. E-mail: c.gaffney@bradford.ac.uk
Copyright # 2008 John Wiley & Sons, Ltd.
into standard blocks (data grids). However, there
are occasions when it is difficult to establish an
accurate grid due to the shape of the survey area
or as a result of obstacles within the area.
Additionally, most commercially available data
collection systems record on a time-basis where
readings are distributed along the length of, or
between predefined markers along, a traverse on
the assumption that the operator has maintained
a constant pace. This can be difficult in some
terrain or when the operator is fatigued. An
Received 15 February 2008
Accepted 7 March 2008
152
extreme example of the errors that can be
introduced in difficult terrain using a fluxgate
magnetometer can be seen in Odah et al. (2005).
The Cyrenaica Archaeological Project is an
international mission under the direction of
Professor Susan Kane (Oberlin College). The
project has a primary focus on the study of
Cyrene, a designated UNESCO World Heritage
site in eastern modern Libya, which was the
leading city of the Libyan Pentapolis. Settled by
Greek colonists toward the end of the seventh
century BC, it remained an active Graeco-Roman
city of distinctively Hellenic character until the
time of the Islamic conquest (AD 643) (http://
www.cyrenaica.org/index.html). As part of the
project it was agreed that a geophysical survey
would be undertaken to provide an additional
context to the many upstanding monuments and
to provide accurate plans of subsurface remains
to support management of the World Heritage
Site as a whole. However, at Cyrene the
abundance of monumental remains, the variation
in topography and the often small areas of open
ground suggested that a traditional 20 or 30 m
grid would be an inefficient way to proceed.
Additionally, potential high temperatures at the
time of the survey (up to 408C) and a highly
variable surface could lead to problems using a
time-based data collection system. As a result it
was decided to consider the Foerster Ferrex #
4.032 system as an alternative. This system is well
known for use in unexploded ordnance detection, although there is an increasing number of
examples of its use for archaeological prospection. In its most basic format data are collected in
the same time-based manner as a traditional
fluxgate system; the separation between the two
fluxgates that comprise each sensing probe is
0.65 m and one or more sensing ‘probes’ may be
attached to a carrying frame.
However, there are additional factors that are
of great interest to those who may consider
surveying using the Foerster instrument.
Initially, although the instrument is based on
fluxgate technology the sensors are ‘factory set’,
and do not require setting up each day or for each
session as with other commercially available
instruments (e.g. Bartington and Chapman,
2004). Secondly, the sensor array is modular;
an array can be built to cover a large swathe
Copyright # 2008 John Wiley & Sons, Ltd.
C. Gaffney et al.
of ground quickly, or the sensors can be
moved close together to create a data-rich survey.
Thirdly, the sensor array can be cart mounted to
reduce the carrying load and to maintain the
sensors at a constant height (Figure 1). Finally,
there is an option to use real time GPS to both
navigate along notional lines and to accurately
locate the real position of the sensor array. This is
done via two GPS data streams and effectively
allows a grid-free survey to be undertaken. In
reality a survey comprises either a single
rectangular data block or a series of rectangular
blocks (Figure 2). The important points to be
made here are that a regular grid is not required
to be physically placed on the surface prior to
data collection and that the data blocks need not
share a common baseline, size or orientation. As a
result awkward shaped areas can be surveyed in
a more efficient manner and the results stitched
together at a later stage using the GPS information associated with the magnetic data.
Inevitably, there are some drawbacks with any
system. Probably the most important is the
reported sensitivity (<0.2 nT), especially when
Figure 1. Photograph of magnetometer in use. This figure is
available in colour online at www.interscience.wiley.com/
journal/arp
Archaeol. Prospect. 15, 151–156 (2008)
DOI: 10.1002/arp
GPS Navigation with Foerster Magnetometer System
153
Figure 2. Schematic diagram ofdata collection.Data collectionwith the Ferrex systemrequires a notionalrectangle thatis defined
bya baseline, thetop cornersof whicharelocatedinthefieldusingthe GPSimmediatelyprior to datacollection.Asaresult the‘best’
location for each survey block can be determined; in practice each survey is made up of several different survey blocks, often on
different orientations. Two strings of information are used from the GPS; one to guide the cart along the centre of the notional
1.5 m swathe, while the second locates the actual position of the cart and, hence, the sensors. Of course the two paths differ due
to obstacles on the surface and operator error.The system automatically collects data on entering the grid and stops on leaving
the notional grid.
one considers the inexorable passage from nano-,
through pico- (Becker, 1995) and on to femtoTesla
(Schultze et al., 2007). A recent paper has
identified this as an important division between
commercially available fluxgate and caesium
vapour instruments (Linford et al., 2007). However, two factors are to be noted here. First, there
is considerable evidence that one of the major
factors in defining small-scale and weak
anomalies is the number of data points over
the target and this is especially true in terms of
the interval between transects (Schmidt and
Marshall, 1997). In the data set described here
the probes were mounted at a distance of 0.5 m
from each other and as readings can be collected
at up to 10 m1 along the traverse the data could
Copyright # 2008 John Wiley & Sons, Ltd.
be regarded as high for archaeological prospecting. A second factor that should be considered is
the type of archaeology to be mapped. In the
current application the archaeology at the site
comprised potential habitation associated with
the core of an ancient town. Experience from
other city sites has shown that magnetometer
responses are usually strong at these sites. Prior
to undertaking the survey at Cyrene a test was
undertaken over a previously surveyed area at
the Roman City of Wroxeter. The results can be
seen in Figure 3 and are highly comparable with
those published previously (Gaffney et al., 2000,
figure 7). Not only are the results equivalent but,
as a consequence of the increased positional
accuracy, there is also a reduction in errors
Archaeol. Prospect. 15, 151–156 (2008)
DOI: 10.1002/arp
C. Gaffney et al.
154
Table 1. Sample statistics from a pilot survey over chalk
subsoil
Track A (forward)
Track B (reverse)
Minimum 5.08
Maximum12.45
Average1.38
Standard deviation 2.79
Median1.1
Minimum 5.39
Maximum11.35
Average1.63
Standard deviation 2.61
Median1.20
for some applications, but such a gross change
will not be acceptable in all cases. As a result code
was written to make an estimate of the ‘background’ for each probe, in both forward and
reverse data collection direction. To illustrate the
probe sensitivity to direction, Table 1 presents
statistics extracted from another pilot survey
over chalk subsoil, where the background
magnetic variation is expected to be low.
These traverses are separated by 0.5 m and were
collected using the same probe. The descriptive
statistics are similar in both cases and suggest
that rotational errors are small.
Sample data from Cyrene
Figure 3. Image of magnetometer data from Wroxeter.
Compare with Gaffney et al. (2000, figure 7).
associated with traditional time-based zig-zag
data collection.
The data shown in Figure 3 have been
de-striped using an algorithm in the Foerster
Dataline software. However, there are limitations
with this algorithm as the shift is only applied as
an integer value. Evidently this may be adequate
Copyright # 2008 John Wiley & Sons, Ltd.
During June 2007 a three-probe cart system was
used to collect magnetometer data in over 40
separate blocks at five locations throughout the
city. A Leica SR530 GPs system was used for the
navigation and after initial trials the system
worked well. One practical point to note is that a
previous upgrade to the Leica firmware meant
that the RTK-fix identifier had been changed; as a
consequence the settings on the Foerster were
different to those identified for a standard SR530
and an alternative software driver was required
to link the two pieces of equipment.
Data collection was relatively straightforward
despite the high temperature and the uneven
surface. Surprisingly, the former was more of a
problem for the operator than the instrument,
with no drift apparent in the data. In some places
the ground surface was particularly uneven and
it is evident that noise levels were relatively high
as a result, although positional accuracy was
consistently good.
Archaeol. Prospect. 15, 151–156 (2008)
DOI: 10.1002/arp
GPS Navigation with Foerster Magnetometer System
155
Figure 4. Magnetometer data (7 nT ¼ black and þ7 nT ¼white) from Cyrene draped over the topographic data for the area.
The large mound is a natural feature with a temple on the top. This figure is available in colour online at www.interscience.wiley.
com/journal/arp
Although it is not possible to record the
elevation data directly onto the Foerster,
we were able to store this information on the
PCMCIA card of the GPS rover using Auto
survey, and to import this into ArcGIS and other
software for data integration. Figure 4 presents a
plot showing data collected in the zone between
the Temple of Zeus and a small hill that has a
temple built on it. Topographic data from
outside the magnetometer data collection areas
have been added to create a local base map. It can
be seen that a street and associated structures
have been identified, along with potential
ordnance fragments associated with World
War II.
ological benefit in equipment that has largely
been used for ordnance detection. Although the
first use of high precision real time GPS for
archaeological geophysics (Leckebusch, 2005)
stressed the increased rapidity of GPR data
collection, the present work demonstrates the
efficiency of magnetometry under relatively
arduous field conditions. The increased measurement density over many other magnetometer
systems and the accurate location of each
measurement point are important counterbalances to the relative insensitivity of the sensors
that are used.
Acknowledgements
Conclusions
From the initial survey using the GPS option for
the Foerster Ferrex # 4.032 system it has been
demonstrated that there is considerable archaeCopyright # 2008 John Wiley & Sons, Ltd.
The authors would like to thank Dr Giuma Anag,
Chairman, Department of Antiquities, and
Abdulkhader el Muzeine, Controller, Department of Antiquities, Cyrene, who have been
extremely helpful and supportive during the
Archaeol. Prospect. 15, 151–156 (2008)
DOI: 10.1002/arp
156
project. CAP would also like to thank all of the
members of the Department of Antiquity and the
Department of Archaeology, University of Omer
al-Mukhtar at el-Beida who took part in the
survey. The team would also like to thank Professor Susan Kane for her continued support
of the project; her experience and knowledge
of the site has proved invaluable during the
two field seasons. The authors would like to
thank Dr Helen Goodchild and Dr Gareth Sears,
co-workers on the project, who facilitated work
in the field and integration with the project GIS
as well as discussing the interpretation and
historical context of the results.
References
Bartington G, Chapman CE. 2004. A high-stability
fluxgate magnetic gradiometer for shallow geophysical survey applications. Archaeological Prospection 11: 19–34.
Becker H. 1995. From nanotesla to picotesla - a new
window for magnetic prospection. Archaeological
Prospection 11: 217–228.
Copyright # 2008 John Wiley & Sons, Ltd.
C. Gaffney et al.
Gaffney CF, Gater JA, Linford P, Gaffney VL, White
R. 2000. Large-scale systematic fluxgate gradiometry at the roman city of Wroxeter. Archaeological Prospection 7: 81–99.
Leckebusch J. 2005. Precision Real-time Positioning
for Fast Geophysical Prospection. Archaeological
Prospection 12: 199–202.
Linford N, Linford P, Martin L, Payne A. 2007.
Recent results from the English Heritage caesium
magnetometer system in comparison with recent
fluxgate gradiometers. Archaeological Prospection
14: 151–166.
Odah H, Abdallatif TF, El-Hemaly IA, Abd El-All E.
2005. Gradiometer survey to locate the ancient
remains distributed to the northeast of the Zoser
Pyramid, Saqqara, Giza, Egypt. Archaeological
Prospection 12: 61–68.
Schmidt A, Marshall A. 1997. Impact of resolution
on the interpretation of archaeological prospection data. In Archaeological Sciences 1995, Sinclair
A, Slater E, Gowlett J (eds). Oxbow Books:
Oxford; 343–348.
Schultze V, Chwala A, Stolz R. 2007. A superconducting quantum interference device system
for geomagnetic archaeometry. Archaeological
Prospection 14: 3.
Archaeol. Prospect. 15, 151–156 (2008)
DOI: 10.1002/arp
Документ
Категория
Без категории
Просмотров
0
Размер файла
289 Кб
Теги
magnetometers, using, site, heritage, foerster, initial, navigation, libyan, system, gps, world, cyren, results
1/--страниц
Пожаловаться на содержимое документа