The Russian research vessel Akademik Nikolai Andreyev participated in the HIFT experiment while anchored above the Krylov Seamount (17.5° N, 30° W). The propagation path length from Heard Island to Krylov was 12.5 Mm, and was located just between the Ascension station (9 Mm) and the more distant American and Canadian stations (16 to 17 Mm). All 35 successful transmissions were received by two omnidirectional hydrophones and a 30-m-long array. Receptions were sampled at 228 Hz and processed in both time and frequency domains. The main result of the processing was quite unexpected: the signal exhibited strong amplitude variability (fading as deep as 20 dB in less than 5 min) and simultaneously exhibited an incredible phase stability. All the phase variance up to the accuracy of positioning was due to the mutual motion of the transmit and receive ships. The minimal propagation loss for cw transmissions was 120 dB. M-sequence processing indicated an 18- to 20-s arrival spread estimate. However, the arrivals were very unstable with a lifetime of about 5 min.

A method is described for measuring the amplitude-frequency and spatial (directional) characteristics of underwater acoustic arrays in real environments, i.e., in the presence of interfering reflections. The method is based on the application of time-delay spectrometry and wideband swept-frequency signals. The results of fullscale experiments are reported, demonstrating the high accuracy, noise immunity, and efficiency of the proposed method.The measurement of the amplitude-frequency response (AFR) characteristics of underwater acoustic arrays in real environments poses a very complex problem. The influence of sound rays reflected from the surface of the water, wall surfaces, the bottom, suspension elements, etc., must be eliminated. Equally important is the assurance of adequate noise immunity so that accurate measurements can be performed in any water region independently of weather conditions.The tone-burst (TB) method is traditionally used for these tasks. However, the measurement accuracy at low frequencies is diminished by the influence of reflecting surfaces. It is generally inappropriate to consider such a thing as noise immunity in relation to the TB method, because the "complexity" (bandwidth-duration product) of the test signal is equal to unity. A measurement technique based on the use of a swept-frequency signal and processing of the signal by time-delay spectrometry (TDS) has been proposed [1] as an alternative. The TDS method has significant advantages not only over the TB method, but also over methods utilizing wideband signals (e.g., a pseudorandom pulse train) with correlation processing. These advantages are summarized as follows:1. The TDS method ensures good arrival-time separation of rays and, as a result, high measurement accuracy near the surface of the water and other reflecting or scattering objects.2. The bandwidth-duration product of the transmitted swept-frequency signals can attain tens of thousands, providing a 40-60 dB or even better improvement in the signal-to-noise ratio (SNR) in comparison with the TB method; the greater dynamic range (more than 80 dB) substantially expands the measurement capabilities and increases the accuracy of the measurements.3. The main signal processing is instrumental (hardware-implemented), permitting measurements to be performed in many channels simultaneously with the use of medium-capacity computers and input equipment.4. The method gives the complete (detailed) AFR over the entire frequency range of the signal after the processing of a single transmission of the sensing signal.5. Modem spectral-analytic methods are capable of improving the accuracy of transit-time measurements a hundredfold over correlation methods, making it possible to monitor the relative positions of elements of the measurement system in the course of the measurements.6. The deterministic and monotonic character of the time variation of the frequency are conducive, first, to "running" (no downtime) measurement of the angular (directional) dependence of the AFR of an array duri...

A tomographic network of seven moored transceivers was deployed in the western Mediterranean basin in early 1994 by IfM (Kiel, Germany), IFREMER (Brest, France), and WHOI (Woods Hole, USA) in the framework of the THETIS-2 project. Signals emitted by the transceivers were also utilized in the 8-week moving ship tomography experiment, MOST. Acoustic measurements were made with a single hydrophone and an eight-element vertical array deployed from a drifting vessel and were complemented by CTD casts. The set of observation points was chosen to provide detailed CTD and acoustic data along existing propagation paths of the THETIS-2 network for comparison of traditional and moving ship tomography under well-controlled environmental conditions as well as to create numerous additional propagation paths to improve resolution in the horizontal plane. By precise compensating the Doppler shift due to vessel drift, the data processing technique used made it possible to resolve eigenray arrivals up to 400–600 km range for different sources. Preliminary results of the MOST data processing, analysis, and interpretation will be presented at the meeting. [Work supported by ISF and INTAS.] a)On leave from P. P. Shirshov Oceanography Institute, Moscow.

A moving ship tomography experiment (MOST) was carried out in the western Mediterranean basin in 1994. Six moored transceivers deployed in the area by IfM (Kiel, Germany), IFREMER (Brest, France), and WHOI (Woods Hole, USA), in the framework of the THETIS-2 project, were utilized as wide-band sound sources for the MOST. An interthermocline eddy with cold core was found during the MOST. Detailed CTD surveys were performed along several sections of the eddy. In the horizontal, it was close to an ellipse with axes of 25 and 40 nm. The sound-speed difference at given depth inside and outside the eddy was up to 2 m/s. Within the eddy core the sound-speed gradient was an order of magnitude less than in surrounding waters. Acoustic signals from the transceivers were recorded at 16 points within and around the eddy in order to study feasibility of reconstruction of mesoscale inhomogeneities of this type by acoustic tomography means. A technique of coherent signal processing used made it possible to resolve multiple eigenray arrivals from four transceivers at each point of observations. Results of mathematical modeling of the acoustic effects of the eddy and their comparison with the experimental data will be reported. [Work supported by ISF and INTAS.] a)On leave from P. P. Shirshov Oceanography Institute, Moscow.

A moving ship tomography experiment was carried out in the Western Mediterranean in 1994. Broadband sound signals were emitted by six moored transceivers deployed by IfM (Kiel, Germany), IFREMER (Brest, France), and WHOI (Woods Hole, USA), in the framework of the THETIS–2 project and recorded at a hydrophone deployed from a drifting research vessel. The acoustic measurements were supplemented with a detailed CTD survey. The data processing technique used made it possible to compensate for the Doppler shift due to vessel drift, measure the channel pulse response and estimate the arrival angles of different rays. The arrival pattern proved to be consistent with numerical predictions using adiabatic normal modes and ray theory. The first results of tomographic inversion for a single vertical slice are presented. The travel times of early raylike arrivals and final cutoffs constitute the data set for inversion. For distances over 300 km, late modal arrivals were resolved, identified and incorporated into the data set. An alternative inversion approach based on matching the overall arrival pattern is discussed and compared with traditional schemes. [Work supported by ISF, INTAS, and RBRF.] a)On leave from P. P. Shirshov Oceanography Institute, Moscow, Russia.