Heard Island Feasibility Test Research Articles

Three-dimensional modeling in global acoustic propagation

The ocean is nearly transparent for acoustic propagation at low frequencies (<100 Hz), leading to the detection of signals (seismic events, volcanoes, and man-made signals) at distances as large as the ocean basin. Historically, basin acoustic modeling has neglected out-of-plane effects and has been performed with the model computed in the range/depth plane for multiple radials following geodesics (Nx2D). Both oceanographic and bathymetric features can lead to out-of-plane effects. In this paper, a summary of computational approaches to this problem will be presented, including vertical-mode, horizontal ray hybrid approaches, and full-3D Parabolic Equation modeling. Out-of-plane effects include refraction and diffraction—which have different effects as well as different approaches to modeling. Experiments where 3D propagation effects were significant will be presented within this context, including Perth-Bermuda (1960), the Heard Island Feasibility Test (1993), and a recent seismic tomography test off the coast of Japan (2015). Three physics mechanisms will be addressed: horizontal deflection due to mesoscale eddies and fronts, reflection from islands (refraction), and diffraction behind bathymetric edges.

ATOC and Other Acoustic Thermometry Observations In New Zealand

A summary of participation of the New Zealand group in the ATOC (Acoustic Thermometry of Ocean Climate) program over a five year period is presented. Transmissions from Heard Island were observed in the Tasman Sea during the Heard Island Feasibility Test in 1991. The California-New Zealand underwater sound path was verified with explosive sources in 1992. Single hydrophone observations were made of transmissions to New Zealand from California from an electrically driven source first suspended beneath a floating platform in 1994 and later placed on the ocean bottom at Pioneer Seamount in 1995. Results from these experiments show that acoustic propagation to ranges of order 10 Mm appears to be characterised by large fluctuations occurring with a time scale of a few minutes.

Time compression of M-sequence transmissions in a very long waveguide with a moving source and receiver

The results of correlation processing of the Heard Island Feasibility Test (HIFT) M-sequence transmissions are presented in this paper. Phase modulated M-sequence signals, with a carrier frequency of 57 Hz, originating from the vicinity of Heard Island (Australia) were received by (Canadian Forces Auxiliary Vessel) CFAV ENDEAVOUR in the Northeast Pacific at a range of 17 400 km. The signals were beamformed, Doppler processed using a novel phase-tracking method, and correlated with replica sequences to yield information about the time-spreading of the multi-path arrivals and path stability. The signal-to-noise ratios (SNR) were very low, so that only a few of the HIFT transmissions could be used to obtain useful correlation results; however, the results indicate that the arrival spread was of the order of 28 s, and that propagation paths were stable for periods of minutes despite the motion of both the towed source and towed receiver.

Hot topics in underwater acoustics

Underwater acoustics is in a period of renaissance with researchers actively pursing studies in many areas from short‐range high‐frequency shallow water acoustics to global low‐frequency acoustics. There has been renewed interest in inhomogeneous sediment acoustics as well as propagation in very shallow water bounded by these sediments. The recent Office of Naval Research ‘‘SRP’’ program studying deep water boundary reverberation has not only been fruitful, but has also laid the foundation for innovative research in the area of boundary scattering in coastal water acoustics. On the other scale, the Heard Island Feasibility Test as described by 18 papers in the October 1994 issue of JASA has re‐awakened interest in the myriad of possibilities of monitoring all scales of the ocean using acoustic methods. In an expanding number of efforts, underwater acoustics is distinguishing itself as developing the most physically based signal processing schemes presently being considered in any field. These and other hot topics will be discussed.

Horizontal ray paths of underwater sound from Heard Island to Ascension Island

To interpret the received transmissions from the Heard Island Feasibility Test, we need to know the propagation paths between the source and the receivers. We have constructed the horizontal ray paths between Heard Island and Ascension Island by using Munk et al. 's equations (Munk et al., J. Phys. Oceanogr., 18, (1988) 1876–98), and present the numerical results in this paper. The ray paths exhibited multipath structure with modes 1 to 4 carrying most of the acoustic energy to Ascension. The calculated travel times for modes 1 to 4 were spread over 10 s and range from 6240 to 6250 s. By taking 6245 ± 5 s, the average, as an estimate of travel time for the first arrival and comparing it with the estimate (6257 s) obtained by signal processing, our numerical result is found to be 12 s less than that obtained by measurement. This travel time discrepancy might be due to the rigid-bottom assumption, the uncertainty in sound speed database, and the effects of ocean currents. The influences of currents on the travel time are qualitatively discussed. It appears that ocean currents can be a major factor affecting the computed travel time.

The Heard Island Feasibility Test

In January 1991, the Heard Island Feasibility Test (HIFT) was carried out to establish the limits of usable, long-range acoustic transmissions. Coded acoustic signals transmitted from a source near Heard Island in the southern Indian Ocean were monitored at 16 sites in the North and South Atlantic, the North and South Pacific, the Indian Ocean, and the Southern Ocean. The question posed by HIFT, whether at such global ranges the signals would permit phase-coherent processing and thus yield favorable signal-to-noise levels, was answered in the affirmative. There was no evidence of distress by the local marine mammal population in response to the acoustic transmissions. HIFT was prerequisite to a program for Acoustic Thermometry of Ocean Climate (ATOC). The principal challenges to such a program are discussed.

Heard Island Feasibility Test: Analysis of Pacific path data obtained with a horizontal line array

Preliminary results from the Heard Island Feasibility Test are presented for data obtained with a towed horizontal line array off the coast of southern California. Signal characteristics examined include signal-to-noise ratios, sound-pressure levels, transmission loss, and arrival bearing angles. For the transmissions analyzed to date the mean sound-pressure level was found to be (63.8±2) dB//1 μPa. The mean transmission loss was found to be (150±3) dB which is significantly greater than that found for receivers in the Atlantic at similar ranges. The mean arrival angle was found to be (214±3)°T and a travel time of approximately 3 h 17 m was measured. Measurements compare favorably with model predictions which indicate that the relatively large transmission loss is due to bottom interactions in the south western Pacific.

Comparison of data and model predictions for Heard Island acoustic transmissions

Propagation paths and signal characteristics from the Heard Island Feasibility Test are modeled using a combination of adiabatic mode theory, coupled mode theory, and parabolic equation (PE) methods. Adiabatic mode theory predicts the horizontal propagation paths (including small grazing angle reflections from bathymetry), group travel times, and estimates of signal loss due to bottom interaction. The PE calculations are performed primarily to illustrate modal coupling near sharp oceanic fronts and bathymetry intruding into the sound channel. They are carried out in depth and range along selected horizontal ray paths. For the Heard Island to Christmas Island path, we carry out a time domain pulse transmission using coupled mode equations. The result gives a pulse envelope in approximate agreement with observations.

A ray variability analysis of sound transmission from Heard Island to California

The influence of mesoscale and seasonal ocean variability on three-dimensional acoustic ray paths from Heard Island through the Southern and Pacific Oceans to the California coast is investigated. For the simulation of these global acoustic ray paths, a three-dimensional Hamiltonian ray-tracing code was used. The input sound-speed fields to the acoustic ray model were interpolated from gridded temperature and salinity output data from the Semtner–Chervin eddy-resolving, global, general circulation model. The ray-tracing results have provided input to experimental planners for receiver placement off the California coast in the Heard Island Feasibility Test and shed light on the variability of the acoustic paths, insonified locations, arrival azimuthal angles, and travel times.

Island diffraction and other three-dimensional effects in global acoustics

The adiabatic mode parabolic equation (PE) is an efficient approach for solving three-dimensional ocean acoustics problems [M. D. Collins, J. Acoust. Soc. Am. 94, 2269–2278 (1993)]. This method is used to solve the outgoing wave equations for the mode coefficients, which depend on range and azimuth. The range- and azimuth-dependent wave numbers appearing in these equations correspond to the eigenvalues. The adiabatic mode PE is applied to solve global-scale problems involving diffraction around islands and other three-dimensional effects. The significance of azimuthal coupling is illustrated by comparing solutions generated both with and without azimuthal coupling. The coupling azimuth solutions predict that the shadow zone behind an island is wider than uncoupled azimuth solutions predict. Since the modal phase velocities increase as ocean depth decreases, this result is consistent with the expected result based on ray theory. Adiabatic mode PE solutions are in qualitative agreement with observations from the Perth–Bermuda experiment [Heaney etal., J. Acoust. Soc. Am. 90, 2586–2594 (1991)] and the Heard Island Feasibility Test [Baggeroer and Munk, Phys. Today 45(9), 22–30 (1992)]. a)Permanent address: Planning Systems, Inc., McLean, VA 22102.

Time compression processing of M sequences from the Heard Island Feasibility Trial

Canada was one of nine participating countries in the Heard Island Feasibility Test (HIFT) and had two towed‐array receivers operating, one in the Pacific and one in the Atlantic. Due to the extremely long propagation paths the Canadian data were found to have relatively low signal‐to‐noise ratios that complicated the M‐sequence time‐compression correlation processing. An analysis was carried out to determine if time‐compression processing of the data was possible with the low signal‐to‐noise ratio data available. A means of removing Doppler variations from the data based on phase tracking the heterodyned baseband signal was developed and applied to several of the stronger HIFT receptions. Useful correlation results were obtained using this technique. Initial estimates of the travel time differences for the dominant acoustic arrivals were obtained from the correlations.

Comparison of data and parabolic equation solutions for long-range pulse propagation

A propagation model that combines the horizontal refraction of the adiabatic mode solution and the accurate field computation of the parabolic equation (PE) method is applicable to long-range problems. In a previous application of this model, qualitative agreement was achieved for narrow-band data from the Heard Island Feasibility Test (HIFT). The broad-band extension of this model has been applied to two long-range propagation paths. Data have been analyzed from a single-slice tomography experiment that involved a 1000-km propagation path in deep water and a source centered at 250 Hz. These data were previously modeled with a WKB solution [Duda etal., J. Acoust. Soc. Am. 92, 939–955 (1992)], which breaks down for the lower modes in the sound channel. These results are in agreement with these data even for the arrivals associated with the lower modes. The Heard Island to Christmas Island propagation path of the HIFT has also been analyzed. Several features along this path cause significant mode coupling, including a sloping bottom near the source, the Antarctic Circumpolar Convergence, and a sea mount. Although there is variability in this data due to fluctuations, the observed dispersion and arrival structure is consistent with these results.

Modeling global-scale three-dimensional sound propagation in the ocean with the parabolic equation method

The parabolic equation (PE) method is efficient for solving range-dependent sound propagation problems, including relatively small-scale three-dimensional problems. Global-scale three-dimensional problems are well beyond current computer resources because the size of the computational grid increases with range. The adiabatic normal mode approximation, which ignores the effects of mode coupling, is applied to reduce problems of this class to a practical size. Each of the complex modal coefficients is a function of longitude and latitude and is determined by solving a two-dimensional PE. The wave number appearing in the PE, which is a function of longitude and latitude, corresponds to the modal eigenvalue. An interesting description of global-scale three-dimensional propagation can be obtained by solving this two-dimensional problem for a small number of modes. This approach is orders of magnitude faster than the three-dimensional PE even when all of the modes are considered. This model should be applicable to analyzing results from global-scale experiments such as the Perth to Bermuda transmissions [J. Acoust. Soc. Am. 71, 51–60 (1982)] and the Heard Island Feasibility Test [Baggeroer and Munk, Phys. Today 45(9), 22–30 (1992)].

Parabolic-equation modeling of global-scale acoustic propagation in the ocean including comparison with Heard Island experimental results

An adiabatic normal-mode model was previously applied to explain the arrival structure of signals propagating between nearly antipodal points [Heaney etal., J. Acoust. Soc. Am. 90, 2586–2594 (1991)]. To investigate the role of mode coupling, an approximate full-wave version of this solution has been implemented based on solving a two-dimensional parabolic equation (PE) along modal ray paths. For 57-Hz propagation from Heard Island to California, a finite-difference solution of the energy-conserving PE is obtained with an Iris workstation in nearly real time (i.e., the computation time is comparable to the propagation time). This solution agrees with data received on a vertical array at a site off California during the Heard Island Feasibility Test (HIFT) [Munk etal., J. Acoust. Soc. Am. 90, 2328–2331 (1991)]. In particular, the PE predicts the experimental result that the received energy is concentrated near mode five. Energy migrates in depth with the sound channel starting near the ocean surface at Heard Island, reaching a maximum depth near the equator, and moving back toward the ocean surface approaching California. Significant mode coupling occurs off the southeast coast of New Zealand due to interaction with the seafloor and the Antarctic Circumpolar Current (ACC). Mode coupling due to the ACC was previously predicted for a shorter HIFT path by Shang etal. [J. Acoust. Soc. Am. 90, 2348 (1991)]. a)Permanent address: Planning Systems, Inc., McLean, VA 22101.

The Heard Island global program: Today Heard Island, tomorrow the world.

The objective of the Heard Island global program is to develop and implement a global acoustic network for the detection of climatic variability in the ocean, using the principle that sound travels faster in warmer water. The first phase was the Heard Island feasibility test, which showed in January 1991 that m-sequence coded acoustic signals transmitted 18 000 km could easily be heard and that their travel times could be measured to a few tens of milliseconds; this is sufficient to detect potential ocean warming against background oceanic fluctuations after about ten years of monitoring. The second phase of the global program (FY92–94) will prepare for the implementation of acoustic sources and receivers at sites around the world. The preparations consist of the development and testing of improved sources and receivers, of a network design that takes into account global models of greenhouse warming and ocean response, and of working with the marine mammal research and protection communities so that the acoustic programs can also provide much-needed information on the response of marine mammals to low-frequency sounds. The third and fourth phases of the global program are the actual implementation of the acoustic network (FY94–97) and the operation of the network for at least a decade (FY95–05). The implementation will first take place in a sparse network configuration for broad ocean monitoring that is global, starting in the Atlantic, the Indian, then the Pacific Ocean, followed by implementation of higher-resolution regional (mapping) networks.

Heard Island Feasibility Test: Long-range sound transmission from Heard Island to Krylov underwater mountain

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.

Heard Island feasibility test: Long-range sound transmission from Heard Island to Krylov underwater mountain.

On 26–31 January 1991, the ‘‘Akademik Nikolai Andreev’’ research vessel anchored at Krylov underwater mountain (17.5° north, 30° west) received the low-frequency signals at a distance of 12.5 thousands kilometers from the source disposed at Heard Island. All transmitted signals were received by two omnidirectional hydrophones at the SOFAR axis and 200-m depth, sampled, and collected on the cartridge tapes by the data acquisition system. All three types of signals−continuous wave 57 Hz, pentaline, and phase coded (M sequence)−have been primarily processed in the time and frequency domains. The purposes of data processing were to determine what the intensity of signals are, the signal/noise (S/N) ratio and temporal variation of the signal phase and magnitude, what is the time structure of the multimode signal, and how time stable is this structure. On most of the transmissions the S/N ratio was about 10–15 dB in 1-Hz band at the carrier frequency 57 Hz, and then it decreased to 0–5 dB on last transmissions. Both temporal variation of Doppler shift and phase fluctuation dominantly depended on the nonuniformity of ship’s motion. GPS data of both ships were compared with the phase variance of some received signals. The distinction between GPS and phase estimations of distance variance is about 10–15 m on an average. The cross-correlation processing of the phase-coded signal enabled one to identify the periodical groups of pulse arrivals that follow with the period of transmitted M sequence. The time stability of multipath arrivals was 5 min and less.