So Sanh Kha Nang AP Dung Cac Mo Hinh Am

A Review of Recent Developments in Underwater

A Review of Recent Developments in Underwater Acoustic Modeling

Acoustic ModelingUnderwater Acoustic Modeling

Acoustical Society of America Seattle WA • 23 27 May 2011

Acoustical Society of America Seattle, WA • 23-27 May 2011Seattle, WA • 26 May 2011

Paul C EtterPaul C. EtterPaper: 4pUW6

INTRODUCTION

• Objectives– Review developments in underwater acoustic modeling over the past eight years

Characterize evolution of the modeling inventory over 32 year period 1979 2011– Characterize evolution of the modeling inventory over 32-year period 1979 - 2011• Surveys conducted at eight-year intervals: 1979, 1987, 1995, 2003, 2011• 2003 survey published in book – provides baseline for 2011 survey

• Review modeling applicationsReview modeling applications– Domains of applicability– Emerging trends

• Identify Modeling Capabilities and Existing Inventoriesy g p g– Basic Acoustic Models

• Propagation Models• Noise Models• Reverberation ModelsReverberation Models

– Sonar Performance Models• Active Sonar Models• Model Operating Systems• Tactical Decision Aids

2

• Tactical Decision Aids

• Provide model-selection guidance

MODELING APPLICATIONSDomains of Applicability

• Domains of Applicability– Arise from assumptions imposed while generating tractable mathematicalArise from assumptions imposed while generating tractable mathematical

solutions from governing physics or empirical data– Restrict applicability of models to specific frequency ranges or problem

geometries– May trade accuracy and computational complexity– Influenced by research versus operational applications

• ResearchC d t d i l b t i t– Conducted in laboratory environments

– Computer time is not a critical factor– Accuracy is important

• Operational• Operational– Conducted in the field– Require rapid execution, often under demanding conditions– Modeling accuracy may be subordinate to processing speedModeling accuracy may be subordinate to processing speed

3

MODELING APPLICATIONSEmerging Trends

• Inverse Methods– geoacoustic and seismo-acoustic inversion– time-reversal acoustics

through the sensor parameter estimation– through-the-sensor parameter estimation– acoustic rain gauges

• Signal Processing– adjoint methods– stochastic resonancestochastic resonance– pulse propagation– clutter environments– vectors and clusters– prediction uncertainty in complex environments– high-frequency acoustics– high frequency acoustics– chaotic and stochastic nonlinear ray dynamics

• Underwater Acoustic Networks– channel models– advances in localization methods (range-based versus range-free schemes)advances in localization methods (range based versus range free schemes)– developments in rapid environmental assessments and new applications for gliders

• Marine-Mammal Endangerment– regulatory initiatives and environmental impacts– rising levels of underwater noise

4

g– role of acidification: H2CO3 (carbonic acid) lowers pH, which lowers acoustic attenuation

– seismic operations and protection of whales

MODELING CAPABILITIES Model Hierarchy

• Underwater acoustics– Development and employment of

acoustical methods toI d t f t• Image underwater features

• Communicate information via the oceanic waveguide

• Measure oceanic properties

Modeling• Modeling– Method for organizing knowledge

accumulated through observation or deduced from underlying principles

– Physical (physics-based) models• Conceptual representation of

the physical processes occurring in the ocean

• Same as analytical modelsMathematical models– Mathematical models

• Empirical models– Based on observations

• Numerical models– Based on mathematical

5

– Based on mathematical representations of the governing physics

MODELING CAPABILITIESPropagation Theory

2

2

22 1

tc ∂Φ∂

=Φ∇ tie ωφ −=Φ

• Frequency-Domain Solutions– Ray theory

022 =+∇ φφ k

( ) ( )zyxiGezyxF ,,,,=φ

k = ω / c

– Normal mode

– Multipath expansion

– Fast field / wavenumber integration

P b li i

( ) ( )rGzF ⋅=φ

( ) ( )θφ– Parabolic equation

• Environmental Range Dependence– Range independent (1D)

( ) ( )rGzrF ⋅= ,,θφ

0

Acoustically Acoustically

0

Acoustically AcousticallyAcoustically AcousticallyRange independent (1D)

f(z)

– Range dependent (2D, 3D)

f(z,r), f(z,r,θ) -200

-150

-100

-50

Wat

er D

epth

(m) Acoustically

ShallowAcoustically

Deep

Hypsometrically Shallow

Hypsometrically Deep-200

-150

-100

-50

Wat

er D

epth

(m) Acoustically

ShallowAcoustically

Deep

Hypsometrically Shallow

Hypsometrically Deep

Acoustically Shallow

Acoustically Deep

Hypsometrically Shallow

Hypsometrically Deep

6

-25010 100 1000 10000 100000

Frequency (Hz)

yp y p-250

10 100 1000 10000 100000

Frequency (Hz)

yp y pyp y p

MODELING CAPABILITIESPropagation Models

Technique CAPARAY PLRAY ACCURAY GRAB LYBIN MPP RAYWAVE FACT RANGER BELLHOP GRASS LYCH Pedersen RP-70FLIRT C h t DELTA HARORAY MEDUSA Pl R SHALFACT

Range Independent Range Dependent

R Th FLIRT Coherent DELTA HARORAY MEDUSA PlaneRay SHALFACT GAMARAY FACTEX HARPO MIMIC PWRC TRIMAIN ICERAY FeyRay HARVEST MPC RAYSON XRAY AP-2/5 MODELAB ORCA ADIAB CPMS NAUTILUS WEDGE BDRM NEMESIS POPP ASERT FELMODE PROLOS WKBZ COMODE NLNM PROTEUS ASTRAL Kanabis PROSIM WRAP DODGE NORMOD3 SHEAR2 CENTRO KRAKEN SHAZAM 3D Ocean

Ray Theory

Normal Mode

FNMSS NORM2L Stickler CMM3D MOATL SNAP / C-SNAP COUPLE MOCTESUMA SWAMP

FAME NEPBR Integrated Mode MULE RAYMODE FFP OASES SAFARI CORE RD-OASES SAFRAN Kutschale FFP Pulse FFP SCOOTER OASES-3D RDOASP

Fast Field or Wavenumber

Integration

Multipath Expansion

MSPFFP RPRESS SPARC RDFFP RDOAST AMPE / CMPE HAPE OS2IFD RMPE 3DTDPACCUB / SPLN / CNP1 HYPER OWWE SNUPE 3DWAPE Corrected PE IFD Wide Angle PAREQ Spectral PE DREP IMP3D PDPE TDPE FDHB3D LOGPE PECan Two-Way PEFEPE MaCh1 PE-FFRAME ULETA

Use Single Environmental Specification

Integration

Parabolic Equation

FEPE MaCh1 PE FFRAME ULETA FEPE-CM MONM3D PESOGEN UNIMOD FEPES MOREPE PE-SSF (UMPE / MMPE) 3DPE (NRL-1) FOR3D NSPE RAM / RAMS / RAMGEO 3DPE (NRL-2)

7

Baseline Reference: P.C. Etter, Underwater Acoustic Modeling and Simulation (Taylor & Francis, 2003, 3rd edition).

New since 2003 baseline

MODELING CAPABILITIESSelection Guidance for Propagation Models

Model type

ApplicationsShallow water Deep water

Low frequency High frequency Low frequency High frequencyRI RD RI RD RI RD RI RD

Ray theory

Normal mode

Multipath expansion

Fast field

Parabolic equation

[Originally adapted from Jensen (1982)]

Low frequency (< 500 Hz) RI: Range-independent environment

High frequency (> 500 Hz) RD: Range-dependent environment

Modeling approach is both applicable (physically) and practical (computationally)

[Originally adapted from Jensen (1982)]

Modeling approach is both applicable (physically) and practical (computationally)

Limitations in accuracy or in speed of execution

Neither applicable or practical

8

. . . Updated Domains of Applicability

MODELING CAPABILITIESNoise Theory

8090

100110• Ambient Noise

– DeterministicSea State: 3

Shipping Level: 5

1020304050607080

dB

Deterministic

– Level

– Directionality

Shipping noise

Shipping Level: 5

010

1 10 100 1000 10000 100000

Frequency (Hz)

– Shipping noise

– Weather noise

• Beam-Noise Statistics• Beam Noise Statistics– Stochastic

• Analytic

Sim lation

ijkijkijk

A

k

n

j

m

iBZS

ij

∑∑∑=== 111• Simulation

– Large-aperture beam noise

– Shipping noise

kji === 111

m = number of routes in the basinn = number of ship typesAij = number of ships of type j on route i

[Adapted from Moll et al. (1979)]

9

Aij number of ships of type j on route iSijk = source intensity of the kth ship of type j on route iZijk = intensity transmission ratio from ship ijk to receiving pointBijk = gain for a plane wave arriving at the array from ship ijk

MODELING CAPABILITIESNoise Models

Ambient Noise Beam-Noise Statistics ANDES AMBENT BBN Shipping Noise ARAMIS BTL CANARY USI Array Noise

Analytic

CNOISE Sonobuoy Noise DANES DANM DINAMO BEAMPL

S S

Simulation

DUNES DSBN FANM NABTAM ISAAC SIAM - I / II Normal Mode Ambient NoiseRANDI I / II / III RANDI - I / II / III

10

Baseline Reference: P.C. Etter, Underwater Acoustic Modeling and Simulation (Taylor & Francis, 2003, 3rd edition).

New since 2003 baseline

MODELING CAPABILITIESReverberation Theory

• Formulations– Cell scattering

Ocean divided into cells• Ocean divided into cells

• Uniformly distributed scatterers

• Sum contributions of all cells

– Point scattering

TargetBistatic

g

• Statistical distribution of scatterers

• Sum echoes from each scatterer

• Source-Receiver Geometry

[Originally adapted from Hodgkiss (1984)]

Source

Bistatic• Source Receiver Geometry– Monostatic - collocated

– Bistatic - separated in range / depth

– Multistatic – multiple sources / receiversp

11

Receiver

MODELING CAPABILITIESReverberation Models

Cell Scattering Point ScatteringMonostatic Bistatic Monostatic Bistatic

C-SNAP-REV ARTEMIS REVGEN BORIS-SSA DOP BAM RITSHPA Under-Ice Reverberation EIGEN / REVERB BiKR SimulationHYREV BiRASP

Cell Scattering Point Scattering

HYREV BiRASP MAM BISAPP PAREQ-REV BISSM PEREV BISTAR PERM-2D OGOPOGO REVMOD RASP REVPA RUMBLE REVSIM S-SCARAB R-SNAP TENAR

12

Baseline Reference: P.C. Etter, Underwater Acoustic Modeling and Simulation (Taylor & Francis, 2003, 3rd edition).

New since 2003 baseline

MODELING CAPABILITIESSonar Performance Theory

• Active Sonar Performance– Stand-alone models

SL = Source Level (of sonar)TL = Transmission Loss (one way)TS = Target StrengthNL = Noise LevelStand alone models

– Solve active sonar equations

RDDINLTSTLSL 2 RDRLTSTLSL 2

DI = Directivity IndexRL = Reverberation LevelRDN = Recognition Differential (noise-limited)RDR = Recognition Differential (reverberation-limited)

• Model Operating Systems

NRDDINLTSTLSL +−=+− 2 RRDRLTSTLSL +=+− 2

– System architectures (executive/bundled)

– Multiple component models

– Data-management software

• Tactical Decision Aids– Operational guidance and training

13

– Visualization techniques

MODELING CAPABILITIESSonar Performance Models

Active RAYMODE MOCASSINALMOST MOC3D

Active Sonar Models

ALMOST MOC3D ASPM MODRAY CASTAR MSASM CONGRATS NISSM - II ESPRESSO SEARAY GASS SONAR HODGSON SST INSIGHT SUPREMO INSTANT SWAMI / DMOSLIRA SWAT LIRA SWAT LORA UAIM MINERAY

CAAM HydroCAM ASPECTModel Operating Systems Tactical Decision Aids

CALYPSO PRISM IMAT CASS SPPS NECTA GSM - Bistatic

14

Baseline Reference: P.C. Etter, Underwater Acoustic Modeling and Simulation (Taylor & Francis, 2003, 3rd edition).

New since 2003 baseline

SUMMARY

• Characterized modeling applications in terms of:– Domains of applicability– Emerging trends

• Identified available models– Current inventory comprises:

Emerging trends

• 128 propagation models• 21 noise models• 28 reverberation models• 35 sonar-performance models

Inventory updated at 8 year intervals:– Inventory updated at 8-year intervals: 1979, 1987, 1995, 2003, 2011

– Approximately 5 models have been added to the inventory each year

zero-intercept ~ 1967

• Provided model selection guidance for research and operational applications

Ob ti M d l t b i di t f R&D i t t

15

• Observation – Models appear to be proxy indicators of R&D investments.