The capability to remove the water column from a sidescan sonar record and at the same time geometrically correct the seabed image such that it is a correct scale representation of the seabed is now an established part of marine hydrographic survey procedures. Most of the manufacturers of sidescan sonar now offer at least one so-called seafloor mapping sonar recorder in their equipment prospectus. The use of the term ‘seafloor mapping’ for the commercially available recorders is, however, a major euphemism since the scale-rectified images produced have no direct or obvious two-dimensional correlation with position. The production of a map still requires a laborious procedure of mosaicing these images using a compromise between best-fit techniques and direct measurement and plotting, or estimation of the towfish position at each data scan. Once this has been achieved it is also still necessary to transfer a positioning at each data scan. Once this has been achieved it is also still necessary to transfer a positioning grid onto the mosaiced data to create a map in the true sense.
A development program was initiated in late 1984 to produce a rectified seafloor image and a UTM grid, generated on the same graphic recorder. The grid was to be annotated and would be printed in its position and orientation directly onto the sonar image from a knowledge of the position of the sidescan sonar towfish at the time of recording.
At an early stage in the planning of this development it became obvious that in order to achieve the primary objective, the production of strips of sonar data which could be laid alongside each other and precisely positioned to form accurate seabed mosaics, it was necessary that the grid generated should be rectilinear. No matter how careful the survey crew may be it is extremely optimistic to expect any sidescan sonar towfish to traverse precisely along straight survey lines; therefore it became necessary to devise a method whereby the sidescan sonar data could be slewed relative to the grid. This immediately restricted the range of seafloor mapping recorders that could be used, both because of technical difficulties in slewing the sonar data and the need for sufficient display space on the recorder to allow some movement without corrupting other data.
The concept of merging sidescan sonar data and position information in this manner is not new, but previous successful attempts have involved extensive computer-aided processing on large office-based mainframe systems, incurring substantial expense and time. The new SONARMAP system operates in real-time using a commercially available sidescan system and Hewlett Packard 9800 series computers, which are in standard commercial use by most marine survey companies.
The ability to create efficiently in real-time a scaled rectilinear grid printed directly over a good geometrically corrected sidescan sonar image of the seabed makes the previously extremely laborious task of patchwork mosaicing of sonar images of areas of seabed very efficient. The superimposition of the grid also makes the exact positioning of features such as seabed pockmarks.
