harvard universityp.f.j. lermusiaux physical-acoustical data assimilation via esse for improved...

38
Harvard University P.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER ocean physics P.F.J. Lermusiaux, UNITES team June 18, 2003 1. ERROR SUBSPACE STATISTICAL ESTIMATION: TWIN EXPERIMENT COUPLED ASSIMILATION OF SOUND-SPEED AND TRANSMISSION LOSS REDUCTION OF BROADBAND TL UNCERTAINTIES 2. NUMERICAL SIMULATIONS OF PRIMER OCEAN PHYSICS INFORMATION AND UNCERTAINTIES RESULTS TO DATE 3. CONCLUSIONS AND UNITES SUMMARY

Upload: ferdinand-gray

Post on 20-Jan-2016

215 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Physical-acoustical data assimilation via ESSE

for improved sonar predictions

and numerical simulations of the PRIMER ocean physics

P.F.J. Lermusiaux, UNITES teamJune 18, 2003

1. ERROR SUBSPACE STATISTICAL ESTIMATION: TWIN EXPERIMENT

– COUPLED ASSIMILATION OF SOUND-SPEED AND TRANSMISSION LOSS

– REDUCTION OF BROADBAND TL UNCERTAINTIES

2. NUMERICAL SIMULATIONS OF PRIMER OCEAN PHYSICS

– INFORMATION AND UNCERTAINTIES

– RESULTS TO DATE

3. CONCLUSIONS AND UNITES SUMMARY

Page 2: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

PHYSICAL-ACOUSTICAL FILTERING IN A SHELFBREAK ENVIRONMENT

Page 3: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Acoustic paths considered (as in Shelfbreak-PRIMER),overlaid on bathymetry.

Page 4: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

PAA PAO

P = POA POO

Coupled Data Assimilation and Uncertainty Initialization:ESSE Interdisciplinary Error Covariances

Physics: xO = [T, S, U, V, W]

Acoustics: xA = [Pressure (p), Phase ()]

x = [xA xO]

cO

P = (x – x t ) ( x – x

t )Tˆ ˆ

ESSE ensemble initialization:- Initialize dominant P based on missing/most uncertain variability in IC- Approach: Multi-variate, 3D, Multi-scale (more than random numbers), see Lermusiaux et al (QJRMS, 2000) and Lermusiaux (JAOT, 2002).

Page 5: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Predicted sound-speed uncertainties Predicted broadband TL uncertainties

Coupled Prediction of Uncertaintiesvia Error Subspace Statistical Estimation (ESSE)

(C1: sound-speed along cross-section 1) (TL1: Transmission loss along cross-section 1)

Page 6: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Page 7: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Coupled ESSE data assimilation of sound-speed and TL data

for a joint estimate of

sound-speed and TL fields

(a)

(d)

(b)

(c)

Page 8: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

(a)

(d)

(b)

(c)

Page 9: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

(a)

(d)

(b)

(c)

Page 10: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Predicted PDF of

broadband TL

Page 11: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

After Assimilation

PDF of broadband

TL

Page 12: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

NUMERICAL SIMULATIONS OF PRIMER OCEAN PHYSICS

Ocean Physics Model and Data Uncertainties

– Bathymetry

– BCs: surface atmospheric forcings, coastal-estuary and open-boundary fluxes

– Initial conditions and ocean physics data

– Parameterized processes: sub-grid-scales, turbulence closures, un-resolved processes

• e.g. tides and internal tides, internal waves and solitons, microstructure and turbulence

– Numerical errors: steep topographies/pressure gradient, non-convergence

Page 13: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Smith and SandwellNOAA soundings combined with

Smith and Sandwell (overlaid with GOM bathymetry)

(predicted topography based on gravity anomaly not well compensated for regions with thick sediments)

Uncertainties in bathymetry (from data differences and statistical model)

Page 14: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER
Page 15: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Baugmarter and Anderson, JGR (1996)

Uncertainties in atmospheric forcings (from buoy-data/3d-model differences)

Page 16: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Three-Hourly Atmospheric Forcings:

Adjusted Eta-29km model, 21 July 1996, 2pm EST

Page 17: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Averaged wind-stress time-series

Page 18: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Page 19: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Page 20: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Page 21: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Page 22: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

SST July 9, 1996 SST July 22, 1996

Page 23: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Ring

Initialization

1. Slope and shelf objective analyses

2. Shelfbreak front feature model

Page 24: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

No Atmos. forcings

With Atmos. forcings

Page 25: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER
Page 26: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER
Page 27: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER
Page 28: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

• Oceans physics/acoustics data assimilation: carried-out as a single multi-scale joint estimation for the first time, using higher-moments to characterize uncertainties

• ESSE nonlinear coupled assimilation recovers fine-scale TL structures (10-100m) and mesoscale ocean physics (10km) from coarse TL data (towed-receiver at 70m depth, one data every 500m) and/or coarse C data (2-3 profiles over 40km)

• Two notable coupled processes:

– Shoreward meander of upper-front leads to less loss in acoustic waveguide (cold pool) on shelf

– Corresponding thickening of thermocline at the front induces phase shifts in ray patterns on the shelf

• Broadband TL uncertainties predicted to be range and depth dependent

• Coupled DA sharpens and homogenizes broadband PDFs

CONCLUSIONS: Coupled ESSE Twin-experiments

Page 29: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

• Shelfbreak Front (SBF) and atmospheric forcing– Intermittent atmospheric forcing impose space and time scales on SBF, controlling

some internal instabilities. They must be accounted for in a complete theory!

– Important consequences for scientific and naval uncertainties

• Other processes found– SBF is stronger where topography is steeper (inflow/outflow in simulation)

– SBF has a tendency to bifurcate at Hudson Canyon

– With warm-core rings in slope-water, this leads to sub-surface northeastward flow

• Numerical mesoscale to sub-mesoscale ocean predictions for acoustic predictions is essential, but substantial progress required, both in data and modeling

• Uncertainties in bathymetry, surface atmospheric forcing, un-resolved processes and ocean data are the ones that matter: they are being modeled

CONCLUSIONS: Numerical Simulations of PRIMER Dynamics

Page 30: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

OASIS* Accomplishments

• Probabilistic performance prediction method presently being evaluated by Navy (SOWG)– ECS Passive Sonar End-to-End System (Uncertainty Scientific

Workshop)– Narrowband Sonar End-to-End System About to Start

• ECS TL azimuthal variability (ASIAEX)• Uncertainty Province Characterization from TL spatial

variability at 3 separate locations in ECS• PRIMER vertical array beamforming fluctuations, signal

and noise

* OASIS Group Includes: Abbot, Dyer, Gedney, Emerson, Shanahan

Page 31: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Peter Cable (BBN)

•Characterized environmentally associated temporal and spatial variability of low frequency active sonar signal-to-interference in a well-behaved littoral environment (ACT I)

•Modeled implications regarding target detection performance

Jim Fulford (NRL)

• Utilized geoacoustic methods either employed by NAVO, or under consideration by NAVO to obtain geoacoustical estimates of means and standard deviations

• These statistics have been employed in forward predictions of the active system performance predictions means and standard deviations in littoral settings

• Suggest that uncertainty of signal excess in reverberation limited regions will be less than uncertainty in noise limited regions.

Page 32: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

CS Chiu (NPS)

• Analyzed dependence of TL fluctuation statistics on signal bandwidth using both (Shelfbreak) PRIMER and ASIAEX (SCS) data.

• Performed model simulation of ASIAEX TL fluctuation statistics, compared modeled statistics to measured statistics, and began studying uncertainty in mean TL prediction.

• Integrated PRIMER SeaSoar and moored data to upgrade daily SSP and TL estimates, and examined the sensitivity of the TL estimate to the resolution of the sound speed estimate.

Page 33: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Tim Duda (WHOI)

•Simulation of propagation through coastal internal waves and fronts

• First-order modulation of the mode-coupling impact of nonlinear internal waves by mesoscale features such as fronts

• Effects of the small- and large-scale features can't be treated independently.

•Quantification of temporal signal variability in SWARM, PRIMER and ASIAEX/SCS experiments

• Scintillation index and correlation time of signal energy correlate well with internal wave statistics.

Page 34: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Glen Gawarkiewicz, Chris Linder WHOI Physical Oceanography- Data

Analysis of shelfbreak frontal responseto wind forcing (Winter PRIMER,Southern MAB data sets)

Analysis of near-bottom temperaturesIn MAB using Lobster trap array

Comparison of PRIMER fields with MODAS

Comparison of ASIAEX mesoscale fields withNRL-Stennis forecast model

Page 35: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Jim Lynch (WHOI)

A) Described mechanisms/statistics of scattering sound in PRIMER due to the internal waves, bathymetry, front, and "foot-of-front". Three publications exist concerning this PRIMER work. (Am author or co-author).

1) Sperry et al. (2003) “Characteristics of acoustic propagation to the eastern vertical line array receiver during the summer 1996 New England Shelfbreak PRIMER experiment.” Submitted to IEEE J. Oceanic Eng’g.

2) Fredericks, Colosi, and Lynch (2003). “Analysis of multipath scintillations observed during the summer 1996 New England shelfbreak PRIMER study.” Submitted to IEEE J. Oceanic En’g.

3) Lynch et al (2003) “Spatial and temporal variations in acoustic propag. characteristics at the New England shelfbreak front.” IEEE J. Oceanic Eng’g., 28(1), 129-150.

B) Described mechanisms/statistics of scattering of sound in the ASIAEX experiment (which can provide a useful comparison to the PRIMER results.) Two publications exist concerning this ASIAEX work. (Am author or co-author).

4) Duda, Lynch, Newhall, Wu, and Chiu (2003). “Fluctuation of 400 Hz sound intensity in the 2001 ASIAEX South China Sea Experiment.”Submitted to IEEE J. Oceani Eng'g.

5) Chiu, Ramp, Miller, Lynch, Duda and Tang (2003). “Acoustic intensity fluctuations induced by South China Sea internal tides and solitons.” Sub. to IEEE J. Oceanic Eng'g.

Page 36: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Page 37: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux

Page 38: Harvard UniversityP.F.J. Lermusiaux Physical-acoustical data assimilation via ESSE for improved sonar predictions and numerical simulations of the PRIMER

Harvard University P.F.J. Lermusiaux