TOPAZ evaluation

1
TOPAZ evaluation

• L. Bertino, F. Counillon, P. Sakov
• Mohn-Sverdrup Center/NERSC

GODAE workshop, Toulouse, June 2009
2
(No Transcript)
3
TOPAZ System overview

• System description
• Validation of TOPAZ
• Data Assimilation

4
Atmospheric Data
Satellite Data SLA, SST, Ice, In Situ Data
Sea-Ice model
Atlantic and Arctic model
EnKF Data assimilation system
Gulf of Mexico model
Eco- system model
Uncertainty estimates
Hindcast studies
User-targeted ocean forecasting
Ocean Primary production
Analyze the ocean circulation, sea-ice and
biogeochemistry. Provide real-time forecasts to
the general public and industrial users
5
The TOPAZ model system

• TOPAZ3 Atlantic and Arctic
• HYCOM EVP sea-ice model
• 11- 16 km horizontal resolution
• 22 hybrid layers
• EnKF
• 100 members
• Observations
• Sea Level Anomalies (CLS)
• Sea Surface Temperatures (NOAA)
• Sea Ice Concentr. (AMSR, NSIDC)
• Sea ice drift (CERSAT)
• Argo T/S profiles (Coriolis)
• Runs weekly, 10 days forecasts
• ECMWF forcing
• http//topaz.nersc.no/thredds
• http//thredds.met.no (MERSEA)

6
EnKF Correlations
3rd Jan 2006
8th Nov 2006
7
The HYCOM model

• 3D numerical ocean model
• Hybrid Coordinate Ocean model, HYCOM (U. Miami)
• US Navy global forecasts
• Hybrid coordinate
• Isopycnal in the interior
• Z-coordinate at the surface
• Terrain following (sigma)
• Nesting capability
• Coupled
• Sea-ice model
• Ecosystem models
• Large community (http//www.hycom.org)

8
Nesting

• Bring dynamically consistent information from
  large-scale circulation to coastal seas
• One-way nesting
• Flather condition for barotropic mode
• Avoids reflection of waves at the boundary
• Simple relaxation for the baroclinic mode
• And for the tracers
• Arbitrary resolution and orientation of the
  nested grids

9
Effect of the upgradeWeekly SSS in Dec. 1999,
free run
MICOM BCM
TOPAZ3
TOPAZ4
10
TOPAZ System overview

• System description
• Validation
• Data Assimilation

11
3 Validation criteriacf weather forecasting
(Murphy, 93)

• Consistency
• Are the operational forecasts in agreement with
  known processes of the ocean circulation?
• Accuracy
• How close to reality are the results?
• Performance (value)
• Advantage over any trivial forecast?
• climatology, persistence

12
Validation Metrics

• Problems
• Validating and comparing GODAE systems
  consistently
• Different model horizontal grids / Vertical
  coordinates
• Large amounts of 4D data
• Large data transfers
• Solutions adopted (during Mersea Strand 1,
  2003-2004)
• 4 Classes of output products (3D, 2D, time
  series, residuals)
• Common output grids (1/8th deg, projection...)
• Self-documented file format (NetCDF)
• Inter-operable file access (OPeNDAP/THREDDS)

13
Arctic Metrics
14
Validation against hydrographic data
June07
Sept07
Topaz2
Topaz3
IMR
15
Online comparison to Argo profiles
16
Sparse profiles under iceNPEO deployment 2006
--- TOPAZ NPEO
North Pole Environment Observatory
17
Water fluxes
18
Sea-ice edgeVisual comparison

• Ice concentration from model in color, SSMI 15
  ice contour in black. Ice drift is overlaid.
• Good overall correspondence between model and
  data
• Visual comparison allows identification of
  problematic regions
• West of Novaya Zemlya - a tendency for the ice
  edge to drift too little to the north during a
  forecast cycle
• South of Svalbard (Bear Island) model ice edge
  too far to the north
• Issues related to model physics
• Ice-ocean momentum exchange
• Ice models neglect physics which may be important
  on small scales
• Fast ice
• MIZ

19
Forecast skills by region
Barents Sea
Alaska
Bering Strait
Central Arctic
Greenland Sea
Kara Sea
20
SLA assimilation residualsAzores box
21
MERSEA sections updated

• Blue MERSEA Class2 sections
• Red Sections from IMR

22
TOPAZ System overview

• System description
• Validation
• Data Assimilation

23
Assimilation of Ocean Color in HYCOM-NORWECOM

• Data
• Satellite Ocean Color (SeaWIFS)
• Coupled Model
• HYCOM-NORWECOM
• (7 compartments)
• Problems
• Coupled 3-dimensional
• physical-biological model.
• High-dimension.
• Non-Gaussian variables.
• Perspectives
• Environment monitoring.
• Fisheries.
• Methodological developments
• for future coastal HR systems.

24
Gaussian anamorphosis with the EnKF
Anamorphosis prior transformation of the
variables in a Gaussian space (Bertino et al.
2003) Twin experiments (surface chlorophyll-a
synthetic observations)
Surface CHLa RMS error
EnKF Cut-off of neg. values
Gaussian Anamorphosis EnKF
Simon Bertino (OSD, 2009)