Page 57 - Winter 2020

is an anchor on the seafloor with a wire stretched up to a buoy that sits below the surface to hold the line taut. The sources and hydrophone receivers are mounted on this line. Additional floatation is also mounted on the line to keep the mooring standing upright, but it is subject to ocean currents, so it moves around in a watch circle about the anchor position. An instrument at the top of a 5,000-m mooring could be swept several hundred meters from the latitude and longitude position of the anchor by ocean currents. A LBL array of acoustic transponders, as described in Traditional Underwater Positioning and Local Vehicle Navigation Systems, is typically deployed around each mooring position to track the motion of the sources and receivers throughout the experiment to correct for the changes in distance between the sources and receivers.
Positioning with Long-Range Underwater Acoustic Measurements
The same core concepts of inferring distance from mea-
surements of signal travel time that we see in GNSS and local underwater acoustic networks can also apply at long ranges. Neutrally buoyant oceanographic floats called swallow floats were equipped with acoustic ping- ers to be tracked by a nearby ship; these were adapted to take advantage of the deep sound channel and were subsequently known as SOFAR floats. The first SOFAR float was deployed in 1968 and was detected 846 km away (Rossby and Webb, 1970).
The SOFAR float signals were originally received by the SOund SUrveillance System (SOSUS) of listening stations operated by the US military. This system tracked more than just floats and enemy submarines. It also received acoustic signals from earthquakes, and there is a wonder- ful 43-day record of passively tracking of an individual blue whale, nicknamed Ol’ Blue, as it took 3,200-km tour of the North Atlantic Ocean (Nishimura, 1994).
The existing listening system was convenient, but equipping each float with an acoustic source was tech- nologically challenging and expensive. In the 1980s, the concept was flipped so that the float had the hydrophone receiver, and acoustic sources transmitted to the floats from known locations to estimate range to the float.
The name was also flipped, and the floats are known as RAFOS, an anadrome for SOFAR (Rossby et al., 1986).
RAFOS sources have been useful to track floats in open water, but when there is sea ice present and the float is unable to get to the surface for a GPS position, underwater positioning becomes even more important. A recent study in the Weddell Gyre near Antarctica tracked 22 floats under ice that were unable to surface to obtain position from the GPS for eight months (Chamberlain et al., 2018).
Similar to RAFOS, a separate long-range navigation system in the Arctic used surface buoys to transmit GPS positions to floats and vehicles for under-ice rang- ing, with an accuracy of 40 m over 400-km ranges. This system operated at 900 Hz, with a programmable band- width from 25 to 100 Hz (Freitag et al., 2015).
RAFOS signals have a bandwidth of 1.6 Hz and there- fore less time resolution than a more broadband source. Figure 4, a and b, contrast predictions of the arrival structure at a 1,145-km range for a RAFOS source with a broadband source having a bandwidth of 50 Hz. In both cases, the latest arriving energy is concentrated near the depth of the sound channel axis, corresponding to rays that stayed at depths with low sound speeds. The early arrivals are from rays that ventured into the higher speed regions of the sound speed profile (in Figure 3, dark blue and green) and therefore also span more of the ocean depth. In both cases, we can see that the energy is spread over about 4 s, but the broadband source provides better resolution.
Figure 4, c and d, shows slices of these acoustic predic- tions at a 2,000 m depth. The broadband signal shown in Figure 4d exhibits sharp peaks in the arrival that can be identified with individual ray paths.
The increased bandwidth is one of the design suggestions for a potential joint navigation/thermometry system addressed in Duda et al. (2006). A system of sources is suggested with center frequencies on the order of 100- 200 Hz and a 50-Hz bandwidth.
The acoustic sources used for ocean acoustic tomography applications are broadband sources designed to trans- mit over ocean basin scales. A 2010-2011 ocean acoustic tomography experiment performed in the Philippine Sea featured six acoustic sources in a pentagon arrangement and provided a rich dataset for evaluating long-range
Spring 2021 • Acoustics Today 57