Air/Ocean Coupled Prediction Systems at NRL

MRY

Richard M. HodurNaval Research Laboratory

Monterey, CA

8th HYCOM Workshop19-21 August 2003Camp Springs, MD

Air/Ocean Coupled Prediction Systems at NRL MRY

Outline

•Global Air/Ocean Coupled Modeling:•NOGAPS/POP•Validation of one-way coupled system•Development of two-way coupled system

•Mesoscale Air/Ocean Coupled Modeling:•COAMPS/NCOM•Focus on Mediterranean:

•Adriatic Circulation Experiment•Real-Time Testbed

•Summary/Plans

NOGAPSNavy Operational Global Atmospheric Prediction System

•Complex Data Quality Control•Atmospheric Analysis:

• Multivariate Optimum Interpolation Analysis (MVOI) of Winds and Heights• Univariate Analysis of Moisture

•Ocean Analysis:• 2D Optimum Interpolation Analysis of SST• 3D Ocean MVOI of T, S, SSH, Sea Ice, and Currents

•Nonlinear, Normal Mode Initialization•Hydrostatic, Spectral Atmospheric Model:

• Cumulus Parameterization (Emanuel, MWR 1999)• Shallow Cumulus Parameterization (Tiedtke, ECMWF Report 1984)• PBL Parameterization (Louis, BLM 1982)• Radiation Parameterization (Harshvardhan et. al., JGR 1987)• Convective and Stratiform Cloud Parameterization (Teixeira and Hogan, JC 2002)• Gravity Wave Drag (Palmer et. al., QJRMS 1986)

•Parallel Ocean Program (POP) Ocean Model•Features:

• Over 16,000 Operational Forecasts run at FNMOC• 6 Hour Incremental Data Assimilation Cycle• Current Operational Resolution: T239 (~55 km), 30 Vertical Levels• Approximately 11 minutes/forecast day wall time using 120 O3K processors• Track Forecasts for all Tropical Cyclones w/max wind > 50 knots• Supplies Boundary Conditions to Mesoscale and Wave Models

NOGAPSNOGAPS

Ocean MVOIOcean MVOI Ocean Model(POP)

Ocean Model(POP)

T, S, v

NAVDASNAVDAS

Fluxes, Stresses

SST

T, q, v

Fully Coupled NOGAPSAir-Ocean with Data Assimilation/Forecast Cycle

MVOI: Multivariate Optimum Interpolation AnalysisPOP: Parallel Ocean Prediction Model

NOGAPS/POP Coupled ModelingComparison of Average 5 m Temperature Analysis Error Correction (Top) with

Forecast Model Correction (Bottom) for August 2000

Reduced errors demonstrate importance of model to data assimilation

Analysis-only produces significant errors in coastal

boundary currents

NOGAPS/POP Coupled ModelingNOGAPS/MVOI/POP Results: SST RMS Errors, Forecast vs Persistence

POP: 1/2 degree, 75S-75N, NOGAPS T159 Forcing

SST RMS Forecast Statistics for CY 2002

NOGAPS/Coupled Models Transition StatusComparison of Modeled (Top) and Observed (Bottom) SSH Variability

POP model was able to simulate SSH

variability in many regions. Direct

assimilation of SSH is expected to improve

results.

POP model run without assimilating

SSH

Plots are for calendar year 2002

Coupled Mesoscale Modeling of the Atmosphere and Ocean

Approach•Utilize existing mesoscale atmosphere and ocean data assimilation systems:

•Atmospheric data assimilation system in the Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS™)

•3-dimensional ocean multivariate optimum interpolation analysis (3D OMVOI)

•NRL Coastal Ocean Model (NCOM)

•Initial tests of the coupled system: Focus on the Mediterranean Sea

•Active meteorologically and oceanographically•Navy relevance•Minimizes need for lateral boundary conditions for ocean model

COAMPSTM is a trademark of the Naval Research Laboratory

COAMPS™Coupled Ocean/Atmosphere Mesoscale Prediction System

•Complex Data Quality Control•Analysis:

•Atmosphere: MVOI of u, v, and Heights; Univariate of T, q•Ocean: 2D SST; 3D MVOI of T, S, SSH, Sea Ice, and Currents

•Initialization:•Atmosphere: Hydrostatic Constraint on Analysis Increments, or Digital Filter•Ocean: Stability check

•Model:•Atmosphere:

• Numerics: Nonhydrostatic, Scheme C, Nested Grids, Sigma-z, Flexible Lateral BCs• Parameterizations: PBL, Convection, Explicit Moist Physics, Radiation, Surface Layer

•Ocean: Navy Coastal Ocean Model (NCOM)• Numerics: Hydrostatic, Scheme C, Nested Grids, Hybrid Sigma/z• Parameterizations: Mellor-Yamada 2.5

•Features:• Globally Relocatable (5 Map Projections)• User-Defined Grid Resolutions, Dimensions, and Number of Nested/Parent Grids• Incremental Data Assimilation; Atmosphere - 6 or 12 hours; Ocean - 12 or 24 hours• Applicable for Idealized or Real-Time Applications• Single Configuration Managed System for All Applications

•Operations (Atmospheric Components plus 2D SST Analysis):• FNMOC: 8 Areas, 4 runs/day, grid spacing as low as 6 km, forecasts to 72 hours• Navy Regional Centers: 2 runs/day, grid spacing as low as 3 km, forecasts to 48 hours

COAMPS is a registered trademark of the Naval Research Laboratory

COAMPS™Coupled Ocean/Atmosphere Mesoscale Prediction System

Existing Future

QCPreprocessing

and MVOI Analysis

Initialization and Forecast

QCPreprocessing

and OMVOI Analysis

Initialization and Forecast

Atm

osp

her

eO

cean

COAMPS is a registered trademark of the Naval Research Laboratory

Atmospheric ReanalysesPurpose: Generate high-resolution fields for forcing NCOM

• Cold start at first analysis time• 12 h incremental data assimilation cycle• Hourly output from forecast model• Analysis include SST analysis for each grid

Analysis

COAMPS 24 h Forecast

Analysis

COAMPS 24 h Forecast

Analysis

COAMPS 24 h Forecast

Observations

Observations

Observations

NOGAPS Fields

NOGAPS BC’s

NOGAPS BC’s

NOGAPS BC’s

12h fcst

12h fcst

12h fcst

Result:Hourly surface

forcing fields for extended time

periods

Atmospheric ReanalysesAverage 10 m winds for Oct 98 using Data Assimilation and Cold Starts with 27 km Resolution

Hours 0-11 of all forecasts used

Results indicate

that horizontal resolution

is important to capture gap flow and other surface-forced events

correctly27 km

81 km

10

5

0

Win

d S

pee

d (

m/s

)

Atmospheric ReanalysesAverage 10 m winds for Oct 98 using hourly output from 27 km grid

Hours 0-11 of all forecasts used

Results indicate

thatdata

assimilation reduces

model spin-up as

evidenced by stronger

winds in gap flow

regions and along

coastlines

Data Assimilation

Cold Start

10

5

0

Win

d S

pee

d (

m/s

)

•Atmosphere:•Bora: Strong, localized

northeasterly winds around Istrian peninsula

•Scirocco: Strong, warm southeast winds

•Ocean:• Cyclonic cells in the central

and southern regions

• River runoff and strong winds create large variability in the northern Adriatic

Bora

Po River

Ocean-Atmosphere Nested Modeling of the Adriatic Sea during Winter and Spring 2001

Meteorology and Oceanography in the Adriatic

1. Generate 27 km atmospheric forcing fields over the Med2. Generate 6 km, 2-year spin-up of the Med using forcing from

#1, then 12-hour data assimilation for October 19993. Generate 4 km atmospheric forcing fields over the Adriatic Sea4. Generate 2 km Adriatic forecasts using initial conditions and

inflow from #2, and atmospheric forcing from #3, 1/28/01-6/4/01

Objectives• Simulate Adriatic

atmospheric and oceanic circulation at high resolution

• Document and understand response of the shallow northern Adriatic waters to forcing by the Bora and Po river run-off

• Quantify the effects of coupling (e.g., one-way, two-way, frequency, resolution) on atmosphere and ocean forecasts

• Aid in planning and interpreting Adriatic Circulation Experiment (ACE) observations

6 km NCOM

27 km

81 km COAMPSTM

2 km NCOM

36 km

12 kmCOAMPS

TM

13

2

4

4 km

Momentum, Heat fluxes

Initial conditions and lateral boundary forcing

Mo

men

tum

, H

eat fluxes

Ocean-Atmosphere Nested Modeling of the Adriatic Sea during Winter and Spring 2001

Design of Experiment

Comparison of observed 10 m winds to observations and 25 m ocean current to observations

Comparison of 36 km and 4 km atmospheric winds

Results(1) 4 km and 36

km winds have similar correlation to observations

(2) Ocean model performs better with 4 km winds

Atmosphere

Ocean

(1)

(2)

Results suggest that the consideration of the effects on an ocean model should be a metric in the validation of

atmospheric models and that high-resolution forcing fields improve ocean forecasts

27 km 4 km

Collaboration with Adriatic Circulation Experiment (ACE)

COAMPS™ Fields: 5 October 1999Resolution Comparison: Atmospheric Forcing

Collaboration with Adriatic Circulation Experiment (ACE)2 km NCOM Fields: 5 October 1999

Comparison of Ocean Model Results Using Atmospheric Fields with Different Resolutions

27 km forcing 4 km forcing

Collaboration with Adriatic Circulation Experiment (ACE)Animation of 2 km Adriatic NCOM Simulation for Oct 1999

Hank Perkins, NRL SSC, Version of 06 Oct 2002

Ocean-Atmosphere Nested Modeling of the Adriatic Sea during Winter and Spring 2001

Field Program (2002-2003): Current Measurements

Split

I t a l yAncona

Po

VeniceTrieste

50

100

200NRL, Perkins, 06 Oct 2002

Adige

Piran20 S l o v e n

i a

Paguro Marine Park

x

x

RovinjC r o a t i a

•

••

x

Senigallia

Cesenatico

Max WERArange: 120 km

12

6

8

10

7

SS Lin

e

4

9

5

3

IC

12

KB Line

ND

Lin

e1

2

VR Line

12

3 45

6

CP Line

12

34

56

LEGEND

Center: SEPTR

Center+NRL: ADCP

IRPEM: CM Mooring

IBM: C10 Mooring

OGS: Met + CM Buoy

NIB: Met + CM Buoy

ISDGM: Tower

OGS and U. Hawaii: WERA Radars (3)

IOF: ADCP

OGS: Wave Buoy

OGS: Codars (3)

+ WTG (+ Sal)

IRB: ADCP or RCM

EuroStratiform: ADCP

Real-Time Ocean Data Assimilation/Forecast Test-Bed

Ocean Analysis/Model Components

Real-time ocean data assimilation run on NRL SGI O2K

MVOI, obs from GODAE Server

NCOM (6 km)72h forecast

Cold Start:First-guess: GDEM, Global NCOM, or POPQC observations/MODAS syntheticsMVOI (z-levels)Initialization5-day spin-up

Warm Start:First-guess: NCOM 12 h forecastQC observationsMVOI (z-levels)Increments added to first-guessInitialization

Hourly Atmospheric Stresses and Fluxes (27 km) from

FNMOC COAMPS™ forecast,Lateral Boundary ConditionsFrom Global NCOM or POP

12h 12hNCOM (6 km)72h forecast

MVOI, obs from GODAE Server

SS

T t

o

Atm

os

ph

eri

c

Mo

de

l

SS

T t

o

Atm

os

ph

eri

c

Mo

de

l

Hourly Atmospheric Stresses and Fluxes (27 km) from

FNMOC COAMPS™ forecast,Lateral Boundary ConditionsFrom Global NCOM or POP

COAMPS™/NCOM Web Page

NCOM Sea Surface Temperature12-21 May 2003: Hours 1-12

Climate Guess Analysis Climate Guess Analysis N

All Data Combined -0.277 -0.03 0.003 0.557 0.262 0.101 4133

ERI SHIP -0.699 -0.182 0.113 1.657 0.598 0.426 24Fixed BUOY Temp 0.029 -0.238 0.006 0.422 0.498 0.129 3Drifting BUOY Temp 0.067 -0.159 -0.049 0.386 0.207 0.146 19N14 Night MCSST -0.277 -0.029 0.003 0.545 0.258 0.093 4072Bucket SHIP -0.711 -0.363 0.261 0.761 0.44 0.387 4Hull Sensor SHIP -0.045 0.117 -0.163 0.334 0.436 0.398 11

Mean Bias (C) RMS Error (C)Analysis Verification at 2003020600: SST

Data Type

Sample Validation of SST AnalysisOutput from Ocean MVOI for Real-Time Mediterranean Run at

0000 UTC 6 March 2003

Similar statistics calculated at each analysis time for all variables at all analysis levels

Statistics indicate that OMVOI/NCOM is performing better than climatology or persistence

Preliminary results

suggest that significant differences exist when forcing an

ocean model with 12 h

frequency as opposed to 1

h or 6 h frequency

12 h frequency runs

1 h and 6 h frequency runs

Importance of Temporal Resolution of Ocean Forcing

Comparison of NCOM runs using 1 h, 6 h, and 12 h COAMPS™ forcingComparisons for Gulf of Lion during February 1999

Application of COAMPS™/NCOM in Eastern Pacific

Cold Start Initial Fields and Lateral Boundary Conditions from Global NCOM

SST: tau 0

SST: tau 72

Surface currents: tau 0

Surface currents: tau 72

•Global:•Testing NOGAPS/POP one-way coupled system•Starting two-way coupled tests•Transition to HYCOM for global ocean prediction

•Mesoscale:•Atmospheric Reanalyses

• Importance of horizontal resolution• Importance of data assimilation

•Air-Ocean Coupling• Use unfiltered, native grid fields for ocean forcing• Collaboration with Adriatic Circulation Experiment (ACE)

• New metric for model verification• Importance of horizontal resolution of ocean forcing

• Importance of temporal resolution of ocean forcing• Real-time ocean data assimilation/Forecast test-bed

•Build in HYCOM for Initial/Lateral Boundary Conditions•Two-way coupling

Air/Ocean Coupled Prediction Systems at NRL MRY

Summary/Plans