Precision, accuracy, and application of diver-towed underwater GPS receivers.

Diver-towed global positioning systems (GPS) handhelds have been used for a few years in underwater monitoring studies. We modeled the accuracy of this method using the software KABKURR originally developed by the University of Rostock for fishing and marine engineering. Additionally, three field experiments were conducted to estimate the precision of the method and apply it in the field: (1) an experiment of underwater transects from 5 to 35 m in the Southern Chile fjord region, (2) a transect from 5 to 30 m under extreme climatic conditions in the Antarctic, and (3) an underwater tracking experiment at Lake Ranco, Southern Chile. The coiled cable length in relation to water depth is the main error source besides the signal quality of the GPS under calm weather conditions. The forces used in the model resulted in a displacement of 2.3 m in a depth of 5 m, 3.2 m at a 10-m depth, 4.6 m in a 20-m depth, 5.5 m at a 30-m depth, and 6.8 m in a 40-m depth, when only an additional 0.5 m cable extension was used compared to the water depth. The GPS buoy requires good buoyancy in order to keep its position at the water surface when the diver is trying to minimize any additional cable extension error. The diver has to apply a tensile force for shortening the cable length at the lower cable end. Repeated diving along transect lines from 5 to 35 m resulted only in small deviations independent of water depth indicating the precision of the method for monitoring studies. Routing of given reference points with a Garmin 76CSx handheld placed in an underwater housing resulted in mean deviances less than 6 m at a water depth of 10 m. Thus, we can confirm that diver-towed GPS handhelds give promising results when used for underwater research in shallow water and open a wide field of applicability, but no submeter accuracy is possible due to the different error sources.