Home    Site Map

Research : Underwater Survey Techniques


The aim of this research was to determine the quality of a typical underwater archaeological survey using 3D trilateration with fibreglass tape measures. To achieve this we need to determine the accuracy, precision and reliability of the position of any point in the survey.  The assessment of quality measures the sizes and nature of undetected errors that exist in computed positions.

Three-dimensional trilateration was chosen for this test as the required quality information is produced as a by-product of processing the measurements.  This form of trilateration is often called the 'Direct Survey Method' and was popularised in marine archaeology by Rule (Rule 1989).  The methods used for processing distance measurements is similar to that used by Global Positioning System receivers (UKOOA 1994) and underwater acoustic positioning systems (Kelland 1994).

To calculate the quality metrics we needed a sample data set of multiple sets of tape measurements made between a number of fixed and rigid points on a typical site underwater. No corrections were applied to the measurements for temperature, sag or tension so the results would be close to those achieved on a typical underwater site.  As a suitable site was not available a test site was set up at the base of the Breakwater Fort behind the Breakwater in Plymouth Sound, described in the Case Studies.  This site was used by the Fort Bovisand Underwater Centre as a training ground for commercial divers as it was sheltered, only 10 m deep, had minimal current and had underwater visibility between 2 m and 5 m.

The chosen site contained a number of fixed and rigid structures suitable for recording.  These included two large concrete blocks, the wall of the Fort itself (Fig 1) and a 7 m long ex-Pilot cutter called Tavy.  A network of 21 control points was installed on the structures (Fig 2), the shape was designed to give a large amount of redundancy and minimal sensitivity to depth errors.  The control points installed on the structures were 5mm galvanised coach bolts cemented into pre-drilled holes. The diameter of the control point bolts was accounted for within the processing program.

Over a period of a year many teams of divers had recorded to 1mm resolution a pre-defined set of measurements between the control points.  The same set of tape measures was used for each exercise. As a check for systematic errors, all tape measures were calibrated against a steel tape measure at 5 m, 10 m and 20 m distances and any tape measures with more than 5 mm in error were not used.

To minimise transcription errors, standard recording forms were used and the data was transferred from the form straight into a computer spreadsheet for analysis.  Measurements were exported from the spreadsheet directly into the processing program.

A set of tape measurements was made on a single baseline on land for comparison.  Data from other sites was collected and compared with the results from the test site.


The results of this work are described in the paper below, click on the link to download it:

International Journal of Nautical Archaeology (2003) 32.2

  • Kelland N., 1994, Developments in Integrated Underwater Acoustic Positioning. Hydrographic Journal  71

  • Rule N., 1989, The Direct Survey Method (DSM) of underwater survey, and its application underwater, IJNA 18.2: 157-162

  • UKOOA, 1994, The Use of Differential GPS in Offshore Surveying


Up ] Research Schema ] [ Research Survey ] Research Wordlists ]