Deep Ocean Propagation Research Articles

Deep ocean sound speed profile estimation using matched autoproduct processing

Sound speed profile (SSP) estimation from acoustic measurements alone in ocean environments is often a highly undetermined problem. Many methods have been developed which match arrival times or acoustic pressure observations to estimate a reduced-order SSP described with shape functions or empirical orthogonal functions (EOFs). These EOFs are commonly derived from multiple in situ measurements. This presentation describes an alternative method that matches autoproduct fields to invert for deep ocean SSPs without knowledge of the source waveform. The inverted SSPs are parameterized with EOFs derived from synthetic SSPs extracted from HYCOM (hybrid coordinate ocean model). This method is developed and illustrated using simulated long-range deep-ocean propagation data for frequencies in the 100s of Hz. The technique is also assessed with measured propagation data from the PhilSea10 experiment. [Sponsored by ONR]

ACOUSTIC RAY STABILITY FOR LONG-RANGE SOUND SPEED PROFILE TRANSITION SCENARIOS

Range-dependency in the background sound speed structure for deep ocean propagation scenarios is typical. Here a special case is considered when there is a gradual transition between single and double minimum sound speed profiles. Ray stability, an important measure for many applications, is analyzed. It is done by studying the unperturbed autonomous ray equations in the quasistatic limit. The wave guide is mapped out in terms of constant action curves in the vertical plane of propagation, which reveals a bifurcation effect in the studied wave guides. Conditions for the constancy or a sudden change of the action of a ray trajectory, an effect that impacts on the "use" of a ray, are given.

Development, testing, and applications of site‐specific tsunami inundation models for real‐time forecasting

The study describes the development, testing and applications of site‐specific tsunami inundation models (forecast models) for use in NOAA's tsunami forecast and warning system. The model development process includes sensitivity studies of tsunami wave characteristics in the nearshore and inundation, for a range of model grid setups, resolutions and parameters. To demonstrate the process, four forecast models in Hawaii, at Hilo, Kahului, Honolulu, and Nawiliwili are described. The models were validated with fourteen historical tsunamis and compared with numerical results from reference inundation models of higher resolution. The accuracy of the modeled maximum wave height is greater than 80% when the observation is greater than 0.5 m; when the observation is below 0.5 m the error is less than 0.3 m. The error of the modeled arrival time of the first peak is within 3% of the travel time. The developed forecast models were further applied to hazard assessment from simulated magnitude 7.5, 8.2, 8.7 and 9.3 tsunamis based on subduction zone earthquakes in the Pacific. The tsunami hazard assessment study indicates that use of a seismic magnitude alone for a tsunami source assessment is inadequate to achieve such accuracy for tsunami amplitude forecasts. The forecast models apply local bathymetric and topographic information, and utilize dynamic boundary conditions from the tsunami source function database, to provide site‐ and event‐specific coastal predictions. Only by combining a Deep‐ocean Assessment and Reporting of Tsunami‐constrained tsunami magnitude with site‐specific high‐resolution models can the forecasts completely cover the evolution of earthquake‐generated tsunami waves: generation, deep ocean propagation, and coastal inundation. Wavelet analysis of the tsunami waves suggests the coastal tsunami frequency responses at different sites are dominated by the local bathymetry, yet they can be partially related to the locations of the tsunami sources. The study also demonstrates the nonlinearity between offshore and nearshore maximum wave amplitudes.

Internal wave strength inversion based on the shape of the axial finale.

Flatte had a significant impact on my dissertation and early education in the field of ocean acoustic propagation. In particular, Flatte’s work on the impact of internal wave scattering on deep ocean propagation was central to my early work in adiabatic mode scattering in the presence of deep water internal waves. In this paper, we will revisit some of Flatte’s wave propagation in a random media theory and show its phenomenological impact on deep water propagation, primarily within his explanation of the Slice89 experiment. In this experiment the axial finale (latest arriving energy) is broadened in depth. Flatte correctly asserted that this was due to internal wave scattering to higher modes (with larger depth extents) and used this as a method for heuristically determining the internal wave strength. Results from Heaney [Ph.D. thesis, SIO‐UCSD (1997)] will be presented using the ATOC experiment. More recent results using NPAL data as well as more recent modeling will be presented. The impact of these results on deep water anti‐submarine warfare and the extension of these results to basin scale propagation (>10000‐km ranges) will be made.

Mode coupling induced by random ocean sound speed structure: Mean intensity and revisiting the evolution equations for the cross‐mode coherences.

The seminal papers of Dozier and Tappert (1978), Dozier (1983), and Creamer (1996) have analyzed mode coupling induced by a random ocean sound speed structure to derive transport equations for the range evolution of mean mode energy. A major assumption in this work is that the cross‐mode coherences rapidly decay with range and can thus be ignored. Theoretical predictions of mode energy have been compared to a direct Monte Carlo simulation, yielding favorable results [Dozier and Tappert (1978) and Creamer (1996)], so the general accuracy of the transport equations cannot be questioned. However, if the observable of interest is the mean acoustic intensity, one will quickly realize that the cross‐mode coherences do not decay rapidly with range and thus need to be estimated for accurate acoustic predictions; an example from deep ocean propagation will be provided. This paper presents a theoretical framework for computing cross‐mode coherences and demonstrates why the coherences are not important for the evolution of mode energy.

Steps Towards the Implementation of a Tsunami Detection, Warning, Mitigation and Preparedness Program for Southwestern Coastal Areas of Mexico

The highly vulnerable Pacific southwest coast of Mexico has been repeatedly affected by local, regional and remote source tsunamis. Mexico presently has no national tsunami warning system in operation. The implementation of key elements of a National Program on Tsunami Detection, Monitoring, Warning and Mitigation is in progress. For local and regional events detection and monitoring, a prototype of a robust and low cost high frequency sea-level tsunami gauge, sampling every minute and equipped with 24 hours real time transmission to the Internet, was developed and is currently in operation. Statistics allow identification of low, medium and extreme hazard categories of arriving tsunamis. These categories are used as prototypes for computer simulations of coastal flooding. A finite-difference numerical model with linear wave theory for the deep ocean propagation, and shallow water nonlinear one for the near shore and interaction with the coast, and non-fixed boundaries for flooding and recession at the coast, is used. For prevention purposes, tsunami inundation maps for several coastal communities, are being produced in this way. The case of the heavily industrialized port of Lazaro Cardenas, located on the sand shoals of a river delta, is illustrated; including a detailed vulnerability assessment study. For public education on preparedness and awareness, printed material for children and adults has been developed and published. It is intended to extend future coverage of this program to the Mexican Caribbean and Gulf of Mexico coastal areas.

A partial theory of phase coherence loss in deep ocean propagation

In a fluctuating ocean, coherently adding multiple receptions of repeated transmissions of a signal generally results in a processing gain relative to the background noise only for some finite duration owing to a gradual loss of wave‐field phase coherence. In this talk a theoretical framework for understanding some elements of phase coherence loss in transient wave fields in deep ocean environments is presented. The theory predicts that phase coherence loss is pathdependent and very different at short range than at long range. PE simulations of transient wave‐field phase coherence loss in an idealized time‐varying ocean are presented and shown to be in good qualitative agreement with theoretical predictions. [Work supported by ONR.]

Sensitivity of ray travel times.

Ray in a waveguide can be considered as a trajectory of the corresponding Hamiltonian system, which appears to be chaotic in a nonuniform environment. From the experimental and practical viewpoints, the ray travel time is an important characteristic that, in some way, involves an information about the waveguide condition. It is shown that the ray travel time as a function of the initial momentum and propagation range in the unperturbed waveguide displays a scaling law. Some properties of the ray travel time predicted by this law still persist in periodically nonuniform waveguides with chaotic ray trajectories. As examples we consider few models with special attention to the underwater acoustic waveguide. It is demonstrated for a deep ocean propagation model that even under conditions of ray chaos the ray travel time is determined, to a considerable extent, by the coordinates of the ray endpoints and the number of turning points, i.e., by a topology of the ray path. We show how the closeness of travel times for rays with equal numbers of turning points reveals itself in ray travel time dependencies on the starting momentum and on the depth of the observation point. It has been shown that the same effect is associated with the appearance of the gap between travel times of chaotic and regular rays. The manifestation of the stickiness (the presence of such parts in a chaotic trajectory where the latter exhibits an almost regular behavior) in ray travel times is discussed. (c) 2002 American Institute of Physics.

Chaos in underwater acoustics

The problem of predicting sound propagation in range‐dependent ocean environments has been investigated, in which it is supposed that the environment (volume and/or boundary) varies smoothly in range and is exactly known with arbitrary precision. Although this problem as stated is deterministic and not intrinsically stochastic, it has been discovered from numerical and analytical studies of physically realistic examples drawn from deep ocean propagation, shallow‐water propagation, and surface duct propagation, that ray path solutions exhibit “classical chaos,” namely, unpredictable and stochastic behavior. Ray paths are found to have a continuous spectrum characteristic of noisy stochastic processes, and ray paths are found to have an exponentially sensitive dependence on initial conditions and environmental parameters characteristic of chaotic processes. This phenomenon of chaos in underwater acoustics is caused by the exponential proliferation of catastrophes (caustics) due to the loss of control implied by the nonseparability of variables in the eikonal equation. As a consequence, even when the ocean environment is known exactly, there exists a “predictability horizon” that limits the range to which acoustic fields can be predicted.

NUMERICAL SIMULATIONS OF THE 1964 ALASKAN TSUNAMI

Tsunamis are long-period water waves usually generated by earthquakes. They occur predominantly in the Pacific Ocean and can produce massive inundation and destruction. The last major tsunami that seriously impacted the United States was the 1964 Alaskan tsunami. This tsunami was generated in Alaska and caused damage in Alaska, the Hawaiian Islands, and on the west coast of the continental United States. It is the most documented of all tsunamis and the only tsunami for which there is information on the ground motion that generated the tsunami. Numerical models have been used to simulate tsunami generation and propagation. A numerical simulation of the 1964 Alaskan tsunami that included deep ocean propagation was performed by Hwang et al. (1) and a comparison was made with a deep water gage at Wake Island. Houston (2) performed a numerical simulation of this tsunami including both deep ocean and nearshore propagation in the Hawaiian Islands and presented comparisons with tide gage recordings. Although Houston (2) demonstrated the ability of numerical models to propagate tsunamis from source regions to distant shorelines, the Hawaiian Islands considered in his simulation have a very short continental shelf and the conclusions of his study might not be applicable to typical continental areas that have significantly longer continental shelves.

Theory of propagation in fluctuating oceans using the radiation fransport equation

A new theory of low‐frequency deep ocean propagation has been developed based on the radiation transport for randomly fluctuating oceans. Using the Fokker‐Planck approximation and a novel ray‐bundle approximation, we have derived analytic solutions for the means and spreads of arrivals. The mean arrival is predicted to differ significantly from the unscattered arrival, but the spreads are predicted to be greatly suppressed relative to spreads in an unchanneled medium. Thus the “3/2 law” of Chernov becomes a “1/2 law” due to the sound channel itself. Further, a new Λ parameter has been derived which is nonsingular at caustics and saturates at small values at large range. [Work supported by ONR and NSE.]

Simulation of propagation in fluctuating oceans using the radiation transport equation

We have developed a radiation transport code, based on the Monte Carlo numerical method, which applies to low-frequency deep ocean propagation with random oceanic fluctuations. Refraction in the sound channel is done by means of a bi-cubic spline interpolation in the angle-depth coordinates. Scattering is accomplished by appropriately sampling several scattering kernels including those given by Wilson and Tappert [J. Acoust. Soc. Am. 66, 256 (1979)]. Numerical experiments show that the mean positions of arrivals are significantly displaced, but that spreading about the mean is suppressed due to the anisotropy of the fluctuations and especially due to the presence of the sound channel itself. [Work supported by ONR.]

Approximate ray angle diagram

The ray angle diagram presents loci as a function of depth of the angles which selected rays make with the horizontal as sound propagates in the ocean. It provides useful qualitative information about deep ocean propagation and the vertical distribution of signal and ambient noise power. The loci are based on Snell’s law. It is shown that an approximate ray angle diagram (ARAD) can be easily constructed directly from the sound-velocity profile since on an appropriate scale the loci of one minus the cosine of the ray angle versus depth are approximately mirror images of the sound-velocity profile. The approximations are shown to involve negligible errors for cases of interest in underwater acoustics. The use and construction of the ARAD are illustrated with examples. A simple graphical technique is presented for annotating the ARAD with range information from which rays may be plotted without the use of a computer.

Approximate ray angle diagram

The Ray Angle Diagram presents loci as a function of depth of the angles which selected rays make with the horizontal as sound propagates in the ocean. It provides useful qualitative information about deep ocean propagation and the vertical distribution of signal and ambient noise power. The loci are based on Snell's law. It is shown that an Approximate Ray Angle Diagram (ARAD) can be easily constructed directly from the Sound Velocity Profile since on an appropriate scale the loci of one minus the cosine of the ray angle versus depth are approximately mirror images of the Sound Velocity Profile. The approximations are shown to involve negligible errors for cases of interest in underwater acoustics. The use and construction of the ARAD are illustrated with examples. A simple graphical technique is presented for annotating the ARAD with range information from which rays may be plotted without the use of a computer.

A Normal Mode Method of Solution for Deep Ocean Propagation

The sound velocity distribution in the ocean is considered a function of depth only. The ocean model is separated into three layers. The surface and bottom layers are represented by linear profiles while the middle layer is assumed to have a velocity profile described by a polynomial. The bottom layer is unbounded, and therefore only one solution to the wave equation is needed to describe the field in it. The bounded surface and middle layers require two solutions of the wave equation to depict the “outgoing” and “uncoming” waves. The solutions for the linear regions are in terms of Bessel functions whereas the field in the middle region is defined by two infinite series. After application of the boundary conditions, the constants of the solutions can be determined, and a normal mode evaluation of the solution is given. For typical profiles off the coast of Eleuthera, the integral representation of the solution is found to be rapidly convergent. The Gaussian approximate method for numerical evaluation is used for evaluation of the field.

Underwater sound: deep-ocean propagation. Variations of temperature and pressure have great influence on the propagation of sound in the ocean : Frosch R.A., 1964. Science, 146 (3646): 889–894

Underwater sound: deep-ocean propagation. Variations of temperature and pressure have great influence on the propagation of sound in the ocean : Frosch R.A., 1964. Science, 146 (3646): 889–894