Ocean Prediction Systems: Advanced Concepts And

Ocean Prediction Systems:Advanced Concepts and Research IssuesAllan R. RobinsonHarvard UniversityDivision of Engineering and Applied SciencesDepartment of Earth and Planetary Sciences

System Concepts Research Issue Examples Demonstration of ConceptMulti-InstitutionalExperiment off CaliforniaCoast (AOSN-II)Harvard UniversityPatrick J. Haley, Jr.Pierre F.J. LermusiauxWayne G. LeslieX. San LiangOleg LogoutovRucheng TianChing S. Chiu (NPS)Larry Anderson (WHOI)Avijit Gangopadhyay (Umass.-Dartmouth)

Interdisciplinary Ocean Science Today Research underway on coupled physical, biological,chemical, sedimentological, acoustical, opticalprocesses Ocean prediction for science and operationalapplications has now been initiated on basin andregional scales Interdisciplinary processes are now known to occuron multiple interactive scales in space and time withbi-directional feedbacks

System Concept The concept of Ocean Observing and PredictionSystems for field and parameter estimations hasrecently crystallized with three major components An observational network: a suite of platforms andsensors for specific tasks A suite of interdisciplinary dynamical models Data assimilation schemes Systems are modular, based on distributedinformation providing shareable, scalable, flexibleand efficient workflow and management

Interdisciplinary Data Assimilation Data assimilation can contributepowerfully to understanding and modelingphysical-acoustical-biological processesand is essential for ocean field predictionand parameter estimation Model-model, data-data and data-modelcompatibilities are essential and dedicatedinterdisciplinary research is needed

Interdisciplinary Processes - Biological-Physical-Acoustical InteractionsPhysics - DensityBiology –Fluorescence(Phytoplankton)Acoustics –Backscatter(Zooplankton)Almeira-Oran front in Mediterranean SeaFielding et al, JMS, 2001Griffiths et al,Vol 12, THE SEA

Biological-Physical-Acoustical Interactions Distribution of zooplankton is influenced by both animal behavior(diel vertical migration) and the physical environment. Fluorescence coincident with subducted surface waters indicatesthat phytoplankton were drawn down and along isopycnals, bycross-front ageostrophic motion, to depths of 200 m. Sound-scattering layers (SSL) show a layer of zooplanktoncoincident with the drawn-down phytoplankton. Layer persistsduring and despite diel vertical migration. Periodic vertical velocities of 20 m/day, associated with thepropagation of wave-like meanders along the front, have asignificant effect on the vertical distribution of zooplankton acrossthe front despite their ability to migrate at greater speeds.

Coupled Interdisciplinary Data Assimilationx [xA xO xB]Unified interdisciplinary state vectorPhysics: xO [T, S, U, V, W]Biology: xB [Ni, Pi, Zi, Bi, Di, Ci]Acoustics: xA [Pressure (p), Phase (ϕ)]P ε{}(xˆ – x t ) ( xˆ – x t )TPAA PAO PABP POA POO POBPBA PBO PBBCoupled error covariancewith off-diagonal terms

Data Assimilationin Advanced OceanPrediction Systems

HOPS/ESSE SystemHarvard Ocean Prediction System - HOPS

HOPS/ESSE SystemError Subspace Statistical Estimation - ESSE Uncertainty forecasts (with dynamic error subspace, error learning)Ensemble-based (with nonlinear and stochastic primitive eq. model (HOPS)Multivariate, non-homogeneous and non-isotropic Data Assimilation (DA)Consistent DA and adaptive sampling schemes

HOPS/ESSE Long-Term Research GoalTo develop, validate, and demonstrate an advancedrelocatable regional ocean prediction systemfor the real-time ensemble forecasting andsimulation of interdisciplinary multiscale oceanicfields and their associated errors and uncertainties,which incorporates bothautonomous adaptive modeling andautonomous adaptive optimal sampling

ApproachTo achieve regional field estimates as realistic andvalid as possible, an effort is made to acquire andassimilate both remotely sensed and in situ synopticmultiscale data from a variety of sensors andplatforms in real time or for the simulation period,and a combination of historical synoptic data andfeature models are used for system initialization.

Ongoing Research ObjectivesTo extend the HOPS-ESSE assimilation, real-timeforecast and simulation capabilities to a singleinterdisciplinary state vector of ocean physicalacoustical-biological fields.To continue to develop and to demonstrate thecapability of multiscale simulations and forecastsfor shorter space and time scales via multiplespace-time nests (Mini-HOPS), and for longerscales via the nesting of HOPS into other basinscale models.To achieve a multi-model ensemble forecastcapability.

Examples Illustrating Research IssuesGulf StreamCoupled physical-biological dynamics studied via compatiblephysical-biological data assimilationCombined feature model and in situ data assimilation in westernboundary currentLigurian Sea and Portuguese CoastMulti-scale real-time forecasting in two-way nested domains –Mini-HOPS: faster time scales, shorter space scales, submesoscale synopticityNew England Shelfbreak FrontEnd-to-End system concept with uncertainties, e.g. sonar systemCoupled physical-acoustical data assimilation with coupled errorcovariances

Gulf StreamBrazil Current

Feature Model

Day 7TemperaturePhytoplanktonPhysical Assim.PhytoplanktonCoupled Assim.Day 10

Conclusions – Compatible Physical/Biological Assimilation Physical data assimilation only – adjustment of the physical fields leads tomisalignment between physical and biological fronts, causing spurious cross-frontalfluxes and consequently spurious biological responses (e.g. enhanced productivity). Biological data assimilation only – little or no feedback to the physics. Physical andbiological fronts become misaligned, causing spurious cross-frontal fluxes andconsequently spurious biological responses (e.g. enhanced productivity). Six-step method:a)b)c)d)e)f) initial estimation of synoptic physical featuresmelding physical data into these fields to obtain the best real-time estimatesphysical dynamical adjustment to generate vertical velocitiesinitial estimation of mesoscale biological fields based on Physical-biological correlationsmelding biological data into these fields, andbiological dynamical adjustment with frozen physical fields to balance the biologicalfields with each other, the model parameters, and the 3-D physical transports.The generation of these fields is done in “adjustment space”, outside of thesimulation of interest (“simulation space”).

Mini-HOPS Designed to locally solve the problem of accuraterepresentation of sub-mesoscale synopticity Involves rapid real-time assimilation of high-resolution data ina high-resolution model domain nested in a regional model Produces locally more accurate oceanographic field estimatesand short-term forecasts and improves the impact of local fieldhigh-resolution data assimilation Dynamically interpolated and extrapolated high-resolutionfields are assimilated through 2-way nesting into large domainmodelsIn collaboration with Dr. Emanuel Coelho (NATO Undersea Research Centre)

MREA-03 Mini-HOPS Protocol Regional Domain (1km) run at Harvard in a 2-way nestedconfiguration with a super-mini domain.– Super mini has the same resolution (1/3 km) as the mini-HOPSdomains and is collocated with them From the super-mini domain,initial and boundary conditionswere extracted for all 3 miniHOPS domains for the followingday and transmitted to the NRVAlliance. Aboard the NRV Alliance, themini-HOPS domains were runthe following day, with updatedatmospheric forcing andassimilating new data.MREA-03 Domains

Mini-HOPS for MREA-03Prior to experiment, several configurations were tested leading toselection of 2-way nesting with super-mini at Harvard During experiment:– Daily runs of regional and super mini at Harvard– Daily transmission of updated IC/BC fields for mini-HOPSdomains– Mini-HOPS successfully run aboard NRV AllianceMini-HOPS simulation runaboard NRV Alliance in Centralmini-HOPS domain (surfacetemperature and velocity)

Mini-HOPS for MREA-04 Portuguese Hydrographic Office utilizing regional HOPS Daily runs of regional and super mini at Harvard Daily transmission of updated IC/BC fields for mini-HOPS domains toNURC scientists for mini-HOPS runs aboard NRV AllianceRegional Domain1km resolutionSuper Mini Domain1/3 km resolution

Coupled Physical-Acoustical Data AssimilationEnd-to-End System Concept Sonar performance prediction requires end-to-end scientificsystems: ocean physics, bottom geophysics, geo-acoustics,underwater acoustics, sonar systems and signal processing Uncertainties inherent in measurements, models, transfer ofuncertainties among linked components Resultant uncertainty in sonar performance prediction itself Specific applications require the consideration of a variety ofspecific end-to-end systems

Coupled discrete state vector x (from continuous φi)x [xA xO]Physics: xO [T, S, U, V, W]Acoustics: xA [Pressure (p), Phase (ϕ)]Coupled error covarianceP ε{(xˆ – x ) ( xˆ – x ) }tt TP Coupled assimilationx x- PHT [HPHT R]-1 (y-Hx-);x- A priori estimate (for forecast)x A posteriori estimate (after assimilation)PAA PAOPOA POOcO

PRIMER End-to-End ProblemInitial Focus on Passive Sonar ProblemLocation: Shelfbreak PRIMERRegionSeason: July-August 1996Sonar System (Receiver): PassiveTowed ArrayTarget: Simulated UUV (withvariable source level)Frequency Range: 100 to 500 HzGeometries: Receiver operating onthe shelf shallow water;target operating on the shelf slope(deeper water than receiver)

Environmental-Acoustical Uncertainty Estimation and Transfers,Coupled Acoustical-Physical DA and End-to-End Systemsin a Shelfbreak EnvironmentNote thefrontVariabilityat the frontWarm/coldevents oneach sideExtremeevents

Starting with physical environmental data, compute thePredictive Probability Of Detection (PPD) from firstprincipals via broadband Transmission Loss (TL) Novel approach: coupled physical-acoustical data assimilationmethod is used in TL estimation Methodology:– HOPS generates ocean physics predictions– NPS model generates ocean acoustics predictions– 100 member ESSE ensemble generates coupled covariances– Coupled ESSE assimilation of CTD and TL measurements

Shelfbreak-PRIMER Acoustic paths considered, overlaid on bathymetry.Path 1: Source: at 300m, 400 Hz Receiver: VLA at about 40 km range, from 0-80m depths

Coupled Physical-Acoustical Data Assimilation of real TL-CTD data:First Eigenmode of coupled normalized error covariance on Jul 26Sound-speedComponentShift in frontal shape(e.g. meander)andBroadband TLComponentits acoustic TLcounterpart abovethe source and in thecold channel on theshelf

Coupled Physical-Acoustical Data Assimilation of real TL-CTD data:TL measurements affect TL and C everywhere.Receivers(VLA)Source

Determination of PPD (Predictive ProbabilityOf Detection) using SNRE-PDFSystems - based PDF (incorporatesenvironmental and system uncertainty)SNRE Signal-to-Noise RatioEnvironmentally InducedUsed by UNITES to characterize and transfer uncertaintyfrom environment through end-to-end problems

PredictedPDF ofbroadbandTL

AfterAssimilationPDF ofbroadbandTL

Coupled HOPS/ESSE/NPS Physics/Acoustics Assimilation Oceans physics/acoustics data assimilation: carried-out as a singlemulti-scale joint estimation for the first time ESSE nonlinear coupled assimilation recovers fine-scale TLstructures and mesoscale ocean physics from real daily TL dataand CTD data Shifts in the frontal shape (meander, etc.) leads to more/less inacoustic waveguide (cold pool on the shelf) Broadband TL uncertainties predicted to be range and depthdependent Coupled DA sharpens and homogenizes broadband PDFs

Integrated Ocean Observingand Prediction SystemsAOSN IIPlatforms, sensors andintegrative models: HOPSROMS real-timeforecasting and re-analyses

Coastal upwelling system:sustained upwelling – relaxation – re-establishmentMonterey Bay and California Current System August 2003Temperature at 10mM1 WindsTemperature at 150m