Figure 1. An early version of Deep Sound, photographed during a tethered engineering test.

Deep Sound, illustrated in Fig.1, is designed to free-fall from the sea surface to a pre-assigned depth, at which point a burn wire releases a drop weight, allowing the system to return to the surface under buoyancy. The descent and ascent rates are the same at 0.6 m/s, corresponding to a round-trip travel time to the bottom of the Challenger Deep of just over 10 hours. A critically important component of Deep Sound is a Vitrovex glass sphere, comprised of two hemispheres, with an external diameter of 43.2 cm. The sphere houses a pack of lithium-ion batteries, of the type found in modern lap-top computers, and a suite of microprocessor-controlled electronics for data acquisition, data storage, power management and system control. Outside the sphere, several hydrophones (underwater microphones) are arranged in vertical and horizontal configurations, and a conductivity-temperature-depth (CTD) probe returns environmental data from which the speed of sound as a function of depth is computed. Motion sensors continuously monitor the pitch, roll and yaw of the system, allowing advection due to local current flow to be monitored.

All the data acquired during a deployment of Deep Sound are stored on-board in solid-state memory. It is therefore imperative that the drop weight releases, enabling the system to return safely to the surface with the data. A sequence of fail-safe mechanisms has been integrated into the system, designed to ensure that the burn wire is indeed triggered even though the pre-assigned depth may not have been reached. Recovery of the system is facilitated by three antennas: a radio beacon, an Argos (GPS) beacon and a xenon high-intensity strobe light. Once Deep Sound is back on board ship, the data are downloaded via a wireless link or through hard-wire connectors penetrating the sphere. Other penetrators allow the batteries to be recharged without separating the hemispheres.

Each of the hydrophones has a bandwidth of 30 kHz and is calibrated over a substantial depth range. The motion of the hydrophones through the water induces a local turbulent flow field, which in turn creates low-frequency (below a few hundred Hertz) interference in the acoustic recordings. To reduce the effects of this flow noise, each hydrophone is fitted with an open-pore-foam flow shield, which in effect keeps the turbulent pressure fluctuations away from the active surface of the sensor.

 

Figure 2. Depth versus time profile for one of the Deep Sound deployments in the Mariana Trench.

 

In May 2009, Deep Sound was deployed on three occasions in the Philippine Sea, to depths of 5,100, 5,500 and 6,000 meters. Two further deployments were made in November 2009, in the Mariana Trench, both to a depth of 9,000 meters. The system was successfully recovered after all these deployments and complete data sets were downloaded from the onboard memory. Fig. 2 shows an example from the Mariana Trench of a depth-versus-time profile, along with the sound speed profiles that were measured during the descent and ascent; and an ambient noise spectrum taken at a depth of 8,413 meters at the same location is shown in Fig. 3. This noise spectrum is notable for its smoothness over much of the spectral range, which is testimony to the effectiveness of the flow shield on the hydrophone. By combining the outputs from two vertically aligned hydrophones, it is possible to obtain the vertical coherence of the noise, which provides a measure of the directionality of the noise in the vertical. Such information helps in understanding the effect of the sound speed profile on the spatial and temporal properties of the ambient noise field.

 

 

Figure 3. Ambient noise spectrum from a depth of 8,413 m in the Mariana Trench.

 

The ultimate challenge for Deep Sound is, of course, the Challenger Deep at the southern end of the Mariana Trench. To meet the demands of descending to such a great depth, a new version of Deep Sound is currently being developed with an enhanced depth rating and an extended sensor suite, but still retaining a Vitrovex sphere as the pressure vessel housing the electronics. In the late summer of 2010, we plan to deploy the new system to a depth of almost 11,000 meters, to the bottom of the Challenger Deep. If successful, it will return with continuous acoustic and environmental recordings taken from the surface to the bottom of the deepest known part of the Earths oceans. (Research supported by the Office of Naval Research)