• Introduction
• The Surveying Process
• Post-Processing
• The Adjustment Program
• General Results

Introduction

The Surveying Process

At the end of each day the data was downloaded from the diver unit into a computer in a folder designed to contain all the data for that day. HPASSConvert, the program developed by Alec Duncan at CMST, was used to process the raw data coming from the diver unit.  HPASSConvert, in its first version, allowed the operator to enter the average temperature over the whole dive together with a value for the salinity, providing the essential parameters for the program to calculate the velocity of sound in water. With the velocity of sound, together with various fixed parameters, the program then converted the raw data and calculated the distances from the diver unit to each transponder. It also calculated the depth, obtained from a pressure transducer located on the diver unit probe and calculated the temperature obtained form a temperature sensor on the diver unit. The resulting text file contained the following information for each fix: TIME; TEMPERATURE (in °C); DEPTH (in metres); and the six transponder-to-diver-unit DISTANCES (in metres). Associated with each of these measurements (excluding time) was the standard deviation of the measurement and the number of measurements used to obtain the final value (the program rejecting readings that lay outside a predetermined range).

The Pandora site lies on the edge of the continental shelf in a region where there are numerous reef and lagoon areas. With the tidal range, the water is noticeably variable in the current strength and temperature, suggesting that water from the warm lagoon areas and the colder deep ocean were ebbing and flowing in an unpredictable manner through the site. Readings showed a remarkable variation; temperatures varied between 24.3°C and 28.8°C over the period of the expedition and by up to 0.6°C over the course of a dive. These variations suggested that using the average the temperature for the period of the dive was likely to cause significant errors in the distances. As a result the software was modified to use the temperature at the time of the reading to calculate the velocity of sound.

It was known that the Pandora site was subjected to a considerable tidal range. Initially it was thought that the tidal variation would be insignificant during the course of a dive, particularly as the program was dealing with relative depths. However, during the course of the analysis of the data it became apparent that the tide had an influence on the measurements. The tidal predictions for the Pandora site were obtained and used to determine the depth correction over the period of each dive.

Having developed the HPASSConvert program to a state where it was able to deal with the problems of the processing data from the diver unit, the analysis of the data was then assessed. At the time of the 1999 Pandora Expedition the expedition team was using a program called Site Surveyor, a more advanced version of Nick Rule’s Web program. Up to that time we had used Web but were aware of its limitations, particularly in handling large data sets. Previously, HPASSConvert provided output in Web format, and the data was then adjusted using Web and the final results presented in a text report or graphic DXF format.

An approach was made to Peter Holt of 3H Consulting Ltd seeking his assistance with the development of HPASSConvert so that it would be able to import directly into Site Surveyor. Together with Alec Duncan from CMST HPASSConvert was programmed so that it exported data in a CSV format that could then be directly imported into Site Surveyor. In addition the HPASSConvert was modified to take advantage of the layers option in Site Surveyor. This option allowed each days survey to be placed in a unique layer containing the inter-transponder distances in a CONTROL layer and the distance to fixes in the FIX layer.

This enabled the daily inter-transponder distances to be recalculated by turning on all the CONTROL layers and adjusting the inter-transponder distances providing a progressively more refined result while at the same time providing the ability to determine if the transponders had been disturbed. Once the inter-transponder distances had been adjusted, the position of the transponders was locked so that the inter-transponder distances then took no further part in the point location adjustment. The layer option then allowed all but one set of measurements or FIXes, to be adjusted. This considerably speeds the processing while at the same time keeping all the measurements in a fixed data set. The process is shown in the flowchart:

The flow-chart shows the process of adjusting the new inter-transponder distances, then locking the transponder positions and then adjusting the data for the day. This turned out to be an extremely useful process, enabling the manipulation of data on a day-by-day basis.

The final results of the Pandora survey were exported from Site Surveyor as a DXF file and imported into ArcView GIS for final graphic presentation. A total of 178 points were recorded by 758 measurements, using the HPASS system. The final RMS error was 35 mm which is impressive, given that the system was operating over measured distances of up to 45 m.

It is difficult to represent the overall accuracy of the system in a form that can be easily visualised, particularly on a site where the precise size of the objects is not known and where the positions on objects are not clearly defined. Two examples are illustrated here to demonstrate the accuracy. On stern section of the site a 2m grid frame was found lying on the seabed. The HPASS was placed on each of the four corners of the grid as an experiment to test the accuracy of the system; a plot of the grid is given in Figure 6.

Data processing was complex, because of the evolving understanding of the problems of the relatively untried system. If the system were to be used again in on this site then some form of depth monitoring would be advisable. A simple depth logging system would be an ideal solution, the monitor could be attached to the reference transponder and would record the depth during the duration of the operation of the HPASS system. At the end of the operation the unit can be recovered and the depth and time data downloaded, a correction can be applied to the depths recorded by the HPASS system to compensate for the effect of the tide.

In its present form, HPASS presents an interesting alternative solution to conventional underwater surveying techniques. The system has a number of advantages: it is fast; accurate; it can be operated in low visibility and relatively deep sites; the speed of operation, once the transponders have been deployed, is at least as fast as a three-tape trilateration system; it gives and accurate and a more reliable z-co-ordinate than the three-tape system there is a considerable saving on post-dive processing and there is reliable assessment of errors.