87
A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling RSMAS August 19, 2002

A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

  • Upload
    ryu

  • View
    46

  • Download
    0

Embed Size (px)

DESCRIPTION

A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling. A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling. RSMAS August 19, 2002. - PowerPoint PPT Presentation

Citation preview

Page 1: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

A Hybrid Coordinate Ocean Model(HYCOM) For Data-Assimilative

Ocean Modeling

A Hybrid Coordinate Ocean Model(HYCOM) For Data-Assimilative

Ocean Modeling

RSMAS August 19, 2002

Page 2: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

• The partnering/collaborating organizations are the University of Miami/RSMAS, the Naval Research Laboratory, NOAA/AOML, the Los Alamos National Laboratory, NERSC, LEGI, the Service Hydrographique et Océanographique de la Marine (SHOM), NAVOCEANO, Planning Systems Inc., Orbital Image Corp., and the U.S.Coast Guard.

• A multi-institutional effort on the development and evaluation of a data-assimilative hybrid isopycnal-sigma-pressure (generalized coordinate) ocean model (called Hybrid Coordinate Ocean Model or HYCOM.)

Page 3: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

The primary computational goal is the establishment of a global eddy-resolving real-time ocean forecast system with sophisticated data assimilation techniques that can be efficiently executed on massively parallel computers

Page 4: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

1. The ocean model: the HYbrid Coordinate Ocean Model (HYCOM)

2. In-situ and satellite data

3. Data assimilation techniques

Page 5: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

- Rotating and stratified fluids => dominance of lateral over vertical transport.

- Hence, it is traditional in ocean modeling to orient the two horizontal coordinates orthogonal to the local vertical direction as determined by gravity.

- The choice of the vertical coordinate system is the single most important aspect of an ocean model's design (DYNAMO, DAMÉE-NAB).

- The practical issues of representation and parameterization are often directly linked to the vertical coordinate choice (Griffies et al., 2000).

Background

Page 6: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Currently, there are three main vertical coordinates in use, none of which provides universal utility. Hence, many developers have been motivated to pursue research into hybrid approaches.

Page 7: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

The hybrid coordinate is one that is isopycnal in the open, stratified ocean, but smoothly reverts to a terrain-following coordinate in shallow coastal regions, and to pressure coordinate in the mixed layer and/or unstratified seas.

Page 8: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 9: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

The capability of assigning additional coordinate surfaces to the oceanic mixed layer allows for sophisticated closure scheme, such as the K-Profile Parameterization (KPP).

Page 10: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Status of HYCOM Development HYCOM 2.0 (released 3 July 2001)

Scalability via MPI and/or OpenMP (2-1000 cpus) Fortran 90 coding style Single source code, for all machine types Bit-for-bit multi-cpu reproducibility

Nesting

Off-line and one-way Based on enclosing region's archive files

MICOM compatibility

MICOM-like mode Can continue a true MICOM simulation Convert MICOM-like to HYCOM-mode Add layers near the surface

Diagnostics

Layer snapshots and animations Cross-section snapshots and animations Means and variabilities Transport sections

Page 11: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

- Add halos for MPI to automatically support periodic boundaries- Support nested-domain open boundaries- Fully global (Pan-Am grid)- Alternative mixed layer models (G. Halliwell presentation) Mellor-Yamada 2.5 Price-Weller-Pinkel

- Orthogonal curvilinear grids- Single passive tracer- NetCDF output files

To be released in September 2002

Other numerical developments (M. Iskandarani presentation)

Page 12: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

HYCOM Long-term Goals for Operational Ocean Prediction

.08 fully – global ocean prediction system transitioned to NAVO in 2006 ~7 km mid-latitude resolution Include shallow water, minimum depth 10-20 m Bi-polar (PanAm) grid for Arctic FY04-06 DoD HPC Challenge project essential Embedded ice model

Increase to .04 resolution globally and transition to NAVO by

the end of the decade ~3.5 km mid-latitude resolution Good resolution for coastal model boundary conditions globally “Baseline” resolution for shelf regions globally

Input to atmospheric and bio-chemical models

John Kindle has NOPP funding for bio-chemical model coupling to HYCOM

Page 13: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Coordinator: Eric P. Chassignet ([email protected])

Page 14: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

NORTH ATLANTIC BASIN-SCALE SIMULATIONS

Based on the CME, DYNAMO and MICOM experience (1˚, 1/3˚ and 1/12˚, respectively)

- The CME comparison is completed (sigma-theta, sigma2, and sigma2 + thermobaricity) [Chassignet et al., to be submitted]

- 1/3˚ routine (interannual forcing – 40 years) [Hogan]

- 1/12˚ in progress (years 1999-2002 after spin-up) [Hogan, Wallcraft, Chassignet, Hurlburt]

NORTH ATLANTIC REGIONAL SIMULATIONS

- Intra-American (Caribbean) Sea (standard configuration for data assimilation testing) [Townsend, Wallcraft]

Page 15: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

ECMWF forcing

Page 16: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Atlantic Model Configuration

• Horizontal grid: 1/12˚ (1678 x 1609 grid points, 6 km spacing on average)

• 28°S to 70°N (including the Mediterranean Sea)

• 26 vertical coordinate surfaces (σ-theta reference)

• Bathymetry: Quality controlled ETOPO5

• Surface forcing: wind stress, wind speed, heat flux (using bulk formula), E-P + relaxation to cllimatological surface salinity

• River runoff included

• Buffer zone: ~3° band along the northern and southern boundaries with relaxation to monthly climatological T and S (Levitus)

Page 17: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

10 x 16 Equal Ocean Decomposition

Running on Brainerd (ARL) 58,000 CPU hrs/model year on 160 CPUs 770 GB/model year for daily 3-D output MPI parallelization

Page 18: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Forced by ECMWF 10 m reanalysis monthly climatological wind and thermal fluxes, climatological surface salinity and relaxation to MODAS climatology at the northern and southern boundaries (themohaline component)

1/12° HYCOM ATLANTIC SIMULATION

Sea Surface Height Sea Surface Temperature

~7 km resolution at mid-latitudes

Page 19: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 20: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 21: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

18 degree Mode Water formation

Page 22: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

18 degree Mode Water formation

Page 23: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

1/12° Atlantic HYCOM

Deep Western Boundary Current

Page 24: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Dickson and Brown 1994 (JGR)

1/12° North Atlantic HYCOM

layer 20 mean speed

Model transportsum layers 20-26ρ > 27.8 (NADW)

4.149.3613.77

5.210.713.3

Observed transport from current meters

Denmark Straits Overflow Region

Page 25: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Denmark Straits Overflow

Cold fresh water forms over shelfin Nordic Seas and spills over the Denmark Strait

and entrains moresaline Labrador Sea water

Page 26: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

• Relaxation to mass fluxes, interface depths, T, S, and density is prescribed in a finite-width sponge zone

• No distinction is made between inflow and outflow boundaries

• The “well-posed” boundary conditions developed by Browning and Kreiss (1982) are applied to the barotropic mode, i.e., the barotropic pressure and velocities are advected into/out of the domain via characteristics

Page 27: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

North Atlantic, 0.32°North Atlantic, 0.32°

Intra-Americas Sea, 0.08°

HYCOM Nesting Applied to IAS

CI = 0.3° C

CI = 0.08° C

CI = 0.08° C

Jan 16

May 4

May 4

SST

SST

SST

[Townsend, Wallcraft]

Page 28: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

0.32° North Atlantic HYCOM

Jul 14

Jul 14

0.08° Intra-Americas Sea HYCOM

Forced by 1979-1993 Monthly Mean ECMWF Reanalysis 10 m winds andboundary conditions from the 0.32° North Atlantic HYCOM

Meridional Velocity, De Soto Canyon, Gulf of Mexico at 29.6N

Page 29: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

0.32° North Atlantic HYCOM

0.08° Intra-Americas Sea HYCOM

Jan 16

Jan 16

Page 30: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Atlantic regional modeling

- Coastal seas around Florida Bay [Kourafalou]

• A nested approach is proposed for a high resolution HYCOM application around south Florida (large scale– regional scale– coastal scale)

• The regional model will provide boundary conditions for limited area hydrodynamic, ecosystem and water quality models in Florida Bay

• The simulations will be closely linked to ongoing observational studies in Florida Bay and adjacent seas

Page 31: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

90 W 85 W 80 W

20 N

25 N

30 N

100

500

50050

0

2000

2000

2000

F l o r i d a

Y u c a t a n P e n i n s u l a

C u b a

West F lorida S

helf

Florida Straits

Yucatan Channel

Campeche Bank

SantarenChannel

G ulf of M exico

Florida

Keys

Florida

Bay

• Physical transport processes impose strong hydrodynamic links between Florida Bay and the adjacent coastal seas. • Significant transient inputs can reach Florida Bay from remote regions of the Gulf of Mexico.

The seas adjacent to Florida Bay:Gulf of Mexico, West Florida Shelf and Florida Straits(the shaded area marks the regional model domain)

Page 32: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Recent observational studiesprovide evidence of transport processes linking south Florida coastal ecosystems

Page 33: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

A comprehensive, multi-year, interdisciplinary data set is available for model initialization, forcing and validation

Page 34: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Atlantic regional modeling

- SEED [Jacobs]

Slope To Shelf Energetics And Exchange DynamicsSlope To Shelf Energetics And Exchange Dynamics

To understand the mechanisms that transfer properties (energy, mass, momentum, heat, salt, …) across the shelf slope

Focus AreaFocus Area

Gulf of Mexico, Mississippi Bight, west of the DeSoto Canyon

Page 35: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Atlantic regional modeling

- DIADEM/TOPAZ [Evensen]

Implement and validate a model system for hindcast simulations in the Faroe-Shetland channel

Run the system in a multiyear hindcast simulation to produce current statistics which can be used to derive design criteria for optimal rig-selection and design

Page 36: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Atlantic regional modeling

- Hurricane impact on mixed layer properties [Jacob, Shay, Halliwell]

Effects of the entrainment Effects of the entrainment closure on the oceanic mixed closure on the oceanic mixed layer response during a layer response during a tropical cyclone passagetropical cyclone passage

Page 37: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Hycom Q MovieGaspar

MY2.5

KPP

PWP

Page 38: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

PACIFIC BASIN-SCALE SIMULATIONS

- 1/12° in progress [Metzger, Hurlburt]

PACIFIC REGIONAL SIMULATIONS

- 1/8° to 1/32° Japan/East Sea [Hogan, Hurlburt]

- 1/32° East Asian Seas [Hogan]

Page 39: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

PACIFIC MODEL CONFIGURATION

• Horizontal grid: 1/12° ( 2294 x 1362 grid points, 6.5 km spacing on average)• 20°S to 65.8°N• 20 vertical coordinates (σ-theta reference)• Bathymetry: Quality controlled ETOP05• Surface forcing: wind stress, wind speed, heat flux (using bulk formula), E-P + relaxation to climatological SSS• River runoff• Buffer zone: ~3° band along southern and eastern boundary with relaxation to monthly climatological T and S• Closed boundaries along 20°S, in the Indonesian throughflow region and in the Bering Strait

Page 40: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

22 x 13 Equal Area Decomposition (all land tiles discarded)

Running on ------- (MHPCC) 50,000 hrs/model year on 207 CPUs 288 GB/model year for 3-D fields every 3 days MPI parallelization

Page 41: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

1/12° Pacific HYCOM SSH Snapshot – 17 December

Forced with climatological HR winds and ECMWF thermal forcing

Page 42: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

1/12° Pacific HYCOM SSH and SST Snapshot – 17 December

Forced with climatological HR winds and ECMWF thermal forcing

Page 43: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Mean Sea Surface Height1/12° Pacific HYCOM vs. Observations

Qu et al. (2001, JPO)

Page 44: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Velocity Cross-section Across Luzon StraitSb-ADCP data versus 1/12° Pacific HYCOM in the upper 300m

Sb-ADCP data from Liang et al. (DSR Part II, in press)HYCOM is forced with high-frequency HR winds and ECMWF thermal forcing

4.4 Sv

3.3 Sv

120.75°E18.5-22°N

Transport

Page 45: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Velocity Cross-section Along Luzon StraitSb-ADCP data versus 1/12° Pacific HYCOM in the upper 300m

Sb-ADCP data from Liang et al. (DSR Part II, in press)HYCOM is forced with high-frequency HR winds and ECMWF thermal forcing

118.5-124.5°E21.0°N

Page 46: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

1/12° Pacific HYCOM

Impact of Hurricane Julliette(Zamudio, Hurlburt, Metzger)

SST SSH

Page 47: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Plans for JES/EAS Modeling [Hogan, Hurlburt]

• Branching of Tsushima Warm Current• Nearshore Branch Dynamics• Water mass formation (ESIW)• Model-data comparisons• Impact vertical coordinate configuration• Data assimilation

• Ability of HYCOM to robustly simulate shelf (Yellow Sea) and deep (JES) environment• Branching of Tsushima Warm Current from the Kuroshio (where, how, etc.)• Interaction of coastal and large-scale currents

1/32° East Asian Seas (LINKS)

1/32° Japan/East Sea (ONR JES DRI)

Page 48: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Japan/East Sea1/8˚

Surface

Japan/East Sea

1/16˚ Surface

Japan/East Sea

1/16˚Deep

Page 49: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 50: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

1/32° HYCOM East Asian Seas Model

Boundary conditions via one-way nestingand 6 hrly ECMWF 10 m atmospheric forcing

Nested inside 1/8° HYCOM Pacific Basin Model

Page 51: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 52: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 53: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

JES summary

JES HYCOM is running with high horizontal grid resolution and synoptic wind and heat flux forcing

Substantial improvement (especially in separation latitude of western boundary current) as horizontal grid resolution increased from 1/8° to 1/32°

Flow over the shelf along the Nearshore Branch

Winter – barotropic – topographic steering Summer – baroclinic – isopycnal outcropping

HYCOM able to simulate formation and spreading of East Sea Intermediate Water, most appears to be formed in northern JES, but some formed locally near Vladivostok

Page 54: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Arabian Sea model configuration [Baraille]

• (34°E-76°E) ; (10°N-30.3°N)

• 16/20 layers, 1/4° horizontal resolution

• Initial condition extracted from the global MICOM model (Bleck)

• Bimonthly relaxation fields extracted from the global MICOM model

applied at the south boundary, in the Gulf of Aden and in the Gulf of

Oman

• 4-year integration with monthly-averaged ECMWF forcing fields

• Integration since 01/01/2000 with high frequency forcing (6 hours)

• Assimilation of altimetry since 11/01/2001

(mean SSH provided by O. M. Smedstad)

INDIAN OCEAN REGIONAL SIMULATIONS

Page 55: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Density field at 20NSea Surface Temperature

Sea Surface Height

10 day forecast

01/31/2001

Page 56: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

GLOBAL SIMULATIONS

- HYCOM/MICOM comparison (100-year simulation) for quasi-global domain (northern boundary at 65° N) at 2° resolution. [“An oceanic general circulation model framed in hybrid isopycnic-cartesian coordinates” by R. Bleck (available on web site)]

- Fully global in progress: target resolution 1.4° for reanalysis [D. Bi, R. Bleck, A. Wallcraft].

- GODAE intensive period: 1/12° in North Atlantic and North Pacific, .72° elsewhere.

- 1/12° everywhere in 2005.

Page 57: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

HYCOM GLOBAL CONFIGURATION

• A Mercator mesh of resolution 2ºx2º (cos(φ)) is used south of 60ºN. At this latitude, the Mercator projection smoothly transitions to a bipolar projection with poles over Canada and Siberia, but without grid singularity over the ocean area.

Page 58: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

HYCOM GLOBAL CONFIGURATION

Page 59: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

• HYCOM global version (GLBx2.00) represents the ocean by 26 layers vertically, using σ0 (potential density referenced to the sea surface) as vertical coordinate in the ocean interior. However, coordinate surfaces level off to become constant-depth surface wherever the isopycnals they follow outcrop at the sea surface.

This figure shows a snap shot of the layers along the 20ºW meridian.

HYCOM GLOBAL CONFIGURATION

Page 60: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Experimental Design

• Surface forcing: COADS monthly atmospheric climatology, including wind stresses, wind speed, solar radiation, net radiation, moisture, etc.. No fresh water flux (E-P) is applied. In stead, the sea surface salinity is relaxed to Levitus climatology.

• Initialization: the model ocean is initialized with the Levitus climatology, i.e., monthly salinity and temperature.

• Sea ice: Energy load.• Mixed layer: KPP• Integration duration: 38 years.

HYCOM GLOBAL CONFIGURATION

Page 61: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Surface features

• Sea Surface Height (annual mean and seasonal cycle)

• Mixed layer thickness (annual mean and seasonal cycle)

• Surface heat flux (annual mean and seasonal cycle)

• Sea ice concentration (annual mean and seasonal cycle)

HYCOM GLOBAL CONFIGURATION

Page 62: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Water mass properties

• Surface distribution: SST (annual mean and seasonal cycle)SSS (annual mean and seasonal cycle)SSD (annual mean and seasonal cycle)

• Vertical structure (zonal average)Temperature (annual mean and seasonal cycle)Salinity (annual mean and seasonal cycle)Stratification (annual mean and seasonal cycle)

HYCOM GLOBAL CONFIGURATION

Page 63: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Horizontal currents

• Mixed Layer Velocity

Annual mean and Seasonal cycle

• Horzontal currents at 1000m depth

Annual mean and Seasonal cycle

HYCOM GLOBAL CONFIGURATION

Page 64: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

The main source of data is provided by altimetry (see presentation by Thacker/Lee for in-situ data assimilation)

Altimetry gives an estimation of the surface circulation

Page 65: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

- Optimal Interpolation with vertical projection either by using isopycnal modes or Cooper-Haines (NRL, SHOM, RSMAS)

- Ensemble Kalman Filter (EKF) (NERSC)

- Single Evolutive Extended Kalman (SEEK Filter) (LEGI)

- Reduced Order Adaptive Filter [ROAF] (SHOM)

- Reduced Order Information Filter [ROIF] (RSMAS)

Page 66: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

- 1/3° Atlantic version of HYCOM

- Assimilation of the Modular Ocean Data Assimilation System (MODAS) optimal interpolated SSH anomalies from satellite altimetry

- Vertical projection of the surface observations by Cooper-Haines

- Running in near real-time

Present assimilation system

Page 67: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

1/3° Atlantic HYCOM SSH20 November 2000

NO ASSIMILATIONAssimilation of MODAS SSH

analyzed fields

Independent frontal analysis of IR observations performed at theNaval Oceanographic Office overlaid. White line shows the part of the front being observed within the last 4 days. Black line shows the part of the front older than 4 days

Page 68: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

1/3° Atlantic HYCOM SSH30 July 2001

Independent frontal analysis of IR observations performed at theNaval Oceanographic Office overlaid. White line shows the part of the front being observed within the last 4 days. Black line shows the part of the front older than 4 days.

Page 69: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Current European project: TOPAZTTowards an OOperational PPrediction system for the North AAtlantic

and european coastal ZZones

Real-time experiment based on EnKF (NERSC) and SEEK (LEGI)

Sea-surface temperature forecastNovember 21st, 2001

Sea-surface height forecastNovember 21st, 2001

Page 70: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

TOPAZ: project participants

• NERSC: – Coordination – HYCOM model system – EnKF with HYCOM and ecosystem model– Assimilation of ice parameters – Nested model applications

• LEGI:– In situ data assimilation– SEEK with HYCOM and ecosystem model– Impact studies

Page 71: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 72: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 73: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 74: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Validation

Based on current meter moorings

available in Faroe Shetland Channel

Page 75: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Exceedence plot from Faroe Bank Channel

Page 76: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Deliverables

• Total of 9146 grid points

• One hour time resolution

• Velocity, temp., saln., layer depth, wind and SLH

• 32.6 Gigabytes per year

• Database is accessed online on the Internet

• Database with model diagnostics

Page 77: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

Adaptive Filter with HYCOM

• Development of the tangent linear model is finished

• We check the validity of the Taylor formula to determine the window where the approximation is valid, depending on the size of the initial perturbation. This work is completed by the definition of a set of variables (not necessarily the HYCOM variables) which can vary linearily with time. This problem is adressed because hybgen is highly non linear.

• The adiabtic adjoint of HYCOM is under developement.

(development started on 09/17/2001, R. Baraille)

Page 78: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

The Reduced Order Information Filter (ROIF) [Chin and Mariano]

The horizontal (cross-) covariance functions of SSH and velocities in an extended information filter are parameterized assecond-order spatial Markov Random Fields. A vertical projection scheme is used to correct lower-layer thicknesses and velocities.

Page 79: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

How do we define the mean SSH for data assimilation?

Page 80: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 81: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 82: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 83: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 84: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 85: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 86: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling
Page 87: A Hybrid Coordinate Ocean Model (HYCOM) For Data-Assimilative Ocean Modeling

OUTREACH

- Live Access Server (presentation by A. Srinivasan)http://hycom.rsmas.miami.edu/dataServer/

-A web-based ocean current reference site (presentation by A. Mariano and E. Ryan)http://oceancurrents.rsmas.miami.edu