One of the critical elements of underwater exploration and research is reliable surveying technology. It's not just enough use film and video to record what's on the sea floor; for the data to be useful to scientists, every feature has to be plotted precisely in relation to all the others. Only when this is done, and the researchers have developed a comprehensive site plan that includes all the various elements (both man-made and biological), can the project team really begin to understand each shipwreck. As one of the Deep Wrecks Project archaeologists put it, "without good mapping, all you've got is a lot of pretty pictures."

Mapping features on the ocean floor is a complex process that involves sophisticated technology at all levels (see diagram below). Just as it's only been in the last two or three decades that remotely-operated vehicles (ROVs) have been created that can extend humans' reach into the depths, similarly the technology to precisely map sites in the deep ocean have come into being. The two sets of technologies are completely intertwined, with advances in each spurring parallel achievements in the other.

Deep-sea surveying projects often begin by placing transponders around the wreck site. These transponders send a series of signals that, when timed and processed through a computer, allow for very precise recording of features on the site. Placing and removing the transponders is very time-consuming, however, so an alternative method called Ultra-Short Baseline (USBL) positioning has been developed.

The USBL system uses a single sound array mounted on the research vessel to receive and interpret sound waves sent by the ROV on the sea floor. The great advantage of the USBL system is that it requires only a single transducer/hydrophone, and allows precise mapping as soon as the ROV reaches the target area on the bottom. Although USBL receivers are subject to a loss of accuracy due to noise in the underwater environment, this can be offset by the use of directional hydrophones that will only register sounds coming in from a narrow, preset direction (i.e., from the target area). Some UBSL systems have shown good accuracy at depths of up to 6,000m (19,680 feet).

The accuracy of different plotting systems is often expressed as a percentage of the depth to the sea floor. As depths increase, the range of possible error increases and the accuracy of the position decreases. On the current Deep Wrecks Project, the USBL system being used has a theoretical accuracy of one-half of one percent (0.5%, or 0.005). This advertised accuracy has been borne out so far through the first three wrecks visited -- Virginia, Halo and Gulfpenn. So what does 0.5% accuracy actually mean, in practical terms? If a wreck at 1,000m (3,280 feet) depth is being surveyed, the recorded position of the wreck should be within 5 meters (0.5% of 1,000m) of its true position, or about 16 feet. While this is not as precise as the routine plotting used for archaeological sites on land, it is sufficiently precise to allow future researchers to locate the wreck again easily and to do meaningful analyses of the features recorded.