Thousands of measurements are taken in the course of an archaeological excavation and even more when recording historic buildings. It is imperative that these be acquired efficiently and in a format readily accessible to the computers that will be processing the data into CAD models and GIS databases.
The traditional techniques often do not satisfy these requirements and this paper explores the options and compares them in terms of accuracy and speed of acquisition. Some specialized measurement tools are also discussed.
A test field was prepared in the classroom with two points set out four meters apart representing one edge of a trench. A board was placed, at an arbitrary orientation, a few meters from this base line and about a meter below it and the students were asked to determine the location of three points on this board using each of the measurement techniques described below. The results were recorded and converted, using an Excel spreadsheet, to X,Y,Z values which were then compared to the known position of the object. These were also plotted using Autocad to provide a graphic representation of the relative accuracy of the various techniques. Originally the intention was to compare the time required to capture these three points along with their accuracy but this, unfortunately, proved to be impossible.
A tape was laid out along the edge of the "trench" and a second tape was stretched horizontally from it to a point above the object as determined by a plumb bob. The person at the intersection of the two tapes was charged with keeping that intersection at right angles and reading the "X" measurement from the base-line tape while the person handling the other end of the second tape read the "Y" value from its intersection with the string of the plumb bob. When these measurements were completed for all three points the "Z" value was determined by stretching a string equipped with a line level over the point to be measured while the vertical distance was measured with one of the tapes.
Four people were required for these measurements. For the X & Y values; one held the plumb bob, one at each end of the second tape, and one recorded the measurements, though an experienced group might get comparable results with only three by giving the plumb holder the object end of the tape. For the Z values three were needed; one holding the string, one watching the level and taking notes and one making the measurement. If a perfectly vertical measurement were desired a fourth would be needed to handle a plumb bob.
Tape measures were attached to the two nails delineating the balk and the distance from each to a point directly above the object, determined by a person holding a plumb bob, was measured and recorded. The absolute X and Y values were determined using the spreadsheet. The Z value for this group was omitted because the technique for recording it would have been identical to that used for the XYZ method described above.
Here too, four people were required: one handling each tape, one holding the plumb bob, and one scribe. It is not possible to accomplish this with three people by having one person handle both tapes without risking damage to the tapes. The Z value would require three as described above.
Direction, Distance & Depth
This technique is an adaptation of the principle of the plane table. A compass rose was affixed to a corner of the "trench" so that its centre was at the location of one of the nails defining the balk and rotated so that 0ø was aligned with the string stretched between the two nails. Measurements were taken by stretching a tape, held on edge, from the centre point of this rose to a point directly above the object being measured. The measurements recorded were the angle on the rose and the horizontal distance to the object from which the X and Y values were determined using the spreadsheet. The Z value was obtained using a plumb bob which was attached to a ¬" tape measure that had been shortened by the length of the weight, allowing the depth to be read directly.
Typically three people were required for this technique; one on the tape, one on the plumb bob and one reading the rose and taking notes. However with practice, this can easily be reduced to two by combining the tape measure and plumb handler.
It was frequently noted that considerable time was wasted trying to determine the numbers on the tapes, either because they were too small to read or because of confusion between the metric and imperial scales. Several participants also experienced difficulty determining the number of meters for a measurement. These problems are all the result of poorly designed scales and a source of metric-only tapes would greatly enhance the efficiency of all measurement techniques.
Another cause of inefficiency is the universal assumption that the smart end of the tape should be in the trench, over the object. Watching people, even experienced ones, juggling the tapes while taking measurements will clearly show that everything possible should be done to reduce the complexity around the object. The easiest way to do this is to give those charged with fixing the location of the object the dumb end of the tape and to read the numbers at the nail or fixed tape.
Considerable time was spent trying to get the plumb bob to hold still and the use of a rod with a bulls eye bubble would speed things up considerably and, I believe increase accuracy.
A formalized procedure for announcing numbers for the scribe should be agreed upon by all members of the team. If the numbers are taken down in the wrong order bizarre results will result and though only one set of numbers in this test was affected by this error it had a dramatic effect on the calculations until noticed and corrected. It became clear what the problem was only when the points were plotted, a step that might never be taken under normal circumstances. This problem is easily eliminated if the pronouncement of a number is always preceded by its coordinate (ie. "X 203.3" "Y 115.0" "Z 86.6" ).
The following results were obtained:
Maximum Average Standard Method Error Error Deviation X,Y,Z 5.30 cm 2.03 cm 1.43 cm Triangulation 2.46 cm 0.79 cm 0.63 cm D,D,D 8.28 cm 2.24 cm 1.88 cm
Though the results of the measurements using triangulation are clearly superior, especially for an inexperienced crew, all of the measurements were better than expected. There were a few values that were clearly in error but in general, with a little care any of these techniques could be used to give reliable results well within the tolerances required. The question of efficiency while collecting these measurements and the ease of plotting them once gathered, both on paper and for the computer, is probably of more importance.
Tools for DDD Measurements
Some custom tools were created in an effort to facilitate the measuring activity. Brief descriptions of each follow:
This is a CAD drawing of a compass rose, calibrated counter clockwise (the Autocad default) and laminated onto a piece of plywood or plastic. It is intended to be levelled and fastened to the ground at the high corner of each trench and its position and orientation surveyed into the plan of the site. All measurements within the trench would reference this point. A hole at the centre of the rose accepts a pin protruding from a small block of wood that has been fashioned to accept the end of the tape though this should be changed to a bracket for the tape body, for reasons mentioned above.
This is a conventional 3/4" or 1" steel tape to the side of which was been added a bubble level allowing the tape to be held in a more or less level position relative to its blade. In practice this levelling function was of no use, better results being obtained by a visual determination based on the relationship between the tape blade and the rose.
This is a ¬" steel tape attached to a plumb bob and looped to allow for the length of the plumb. It allows the depth of an object to be measured directly, eliminating the need for separate plumb bob and tape handlers.
Sketch Board & Paper
The board is a standard clip board for 8« x 14 paper with a tiny (1mm í 2-3mm long) pin protruding near the upper left or right hand corner (depending on the high corner of the trench) against which a standard architects scale, notched slightly at the zero point, can be rotated. The paper, like the rose was prepared using the CAD program and is intended to allow the draftsperson to sketch directly from the coordinates as they are measured. It has a centre mark which is pierced by the pin and, as a border, on the two sides opposite this mark it is calibrated in the same way as the rose. Coordinates are transferred to these sheets by rotating the scale to the corresponding degree mark and a mark made at the appropriate distance.
Barker, P. 1986
Understanding archaeological excavation.
Cleator, P. 1973
Dever et al. 1978
A Manual of field excavation
Cincinnati, Hebrew Union College.
Joukowsky, M. 1980
A Complete manual of field archaeology
Englewood Cliffs, N.J., Prentice-Hall.
Kenyon, K. 1957
Beginning in Archaeology
New York, Praeger
Nickerson, S. 1994
A Site Information System (SIS):
CADD/Database Integration for Field Use.
APT Bulletin 24(1):56-62
Nickerson, S. 1995
Integration of Measurements and the Knowledge Base
Generating a CAD Model - on Site
Automating the Drafting for as-Found Recording and
Proceedings of ACADIA'95, Seattle Washington
Patias, P. 1994
Desktop Digital Recording and Management Tools:
From Ideas to End Products.
APT Bulletin 24(1):23-27
Ross et al. (ed) 1991
Computing for Archaeologists
Oxford University Committee for Archaeology
Spier, R. 1970
Surveying and mapping; a manual of simplified techniques.
New York, Holt.
Streilein, A. 1993
Application of a S-VHS Camcorder in Architectural
15th International Symposium of CIPA, September 22-25, 1993,
Streilein, A. 1994
Towards automation in architectural photogrammetry:
CAD-based 3D-feature extraction
ISPRS Journal of Photogrammetry and Remote Sensing,
Waldhausl et al. 1988.
Architectural Photogrammetry world-wide and by anybody
using non-metric cameras?
11th International Symposium of CIPA, July 10, 1988,
Sofia Bulgaria. pp. 35-49