1. TOPAZ monitoring and forecasting system
L. Bertino, G. Evensen, K.A. Lisæter, H. Sagen, N. Winther, F. Counillon

2. Objectives
To provide the best possible ocean current forecasts (10 days to 1 month ahead) and improve the understanding of the ocean circulation and variability.
How?
by integrated use of a numerical model and remote sensing via data assimilation.

3. The ingredients 3D numerical ocean model Observations
HYCOM (U. Miami, USA)
Sea Ice model
Biogeoch. model (AWI, D)
Observations
Altimetry, SST (CLS, F)
Sea Ice (NSIDC, USA)
In-situ (CORIOLIS, F)
Data assimilation
Ensemble Kalman Filter [Evensen 1994, 2003, 2004]

4. The TOPAZ model system TOPAZ: Atlantic and Arctic Nested models
18-35 km resolution
22 hybrid layers
EnKF assimilation
Nested models
Arbitrary resolution
Arbitrary orientation
Specific assimilation

5. Weekly Forecast Cycle Weekly cycle
Next month
Forecast
Last week
Present week
Forecast
Analysis
Nowcast
Analysis
Nowcast
Weekly cycle
Atmospheric forcing fields from ECMWF

6. What is data assimilation? Goal: make models respect observations
Optimal control theory
Optimize control parameters (forcing, initial)
To fit measurements
Spatial statistics theory
Measurements have a domain of influence
Physical variables are linked in complex ways
Complementary views on the problem.

7. TOPAZ forecast 23rd Oct. Ocean temperature and ice

8. HYCOM v1.3
ETOPO5 bathymetry
HYCOM v2.1
GEBCO bathymetry

9. HYCOM1
ETOPO5
Feroe -Shetland Ridge
HYCOM2
GEBCO

10. Next efforts Oceanographic validation of the system (1 PhD)
More data to assimilate
Ice thickness
Ice drift
Sea Surface Salinity

11. High-resolution forecasting
1 Post-Doc
1 PhD

12. Gulf of Mexico forecast Oct. 20
Illustration: François Counillon

13. Note Hurricane “Ivan” at day Analysis + 6

14. Validation overlay with ocean color data from MODIS (not assimilated)
From June to Sept. 2004: eddy shedding

15. Future perspectives Nordic Seas (1 Post-Doc + 1 PhD)
Illustrations by H. Sagen and N. Winther

16. And more …
Illustr. G. Evensen
Illustr. C. Hansen

17. Ecosystem forecasting
1 Post Doc
1 PhD student

18. Ecosystem modelling 3D ecosystem model: FDM coupled with HYCOM
Illustration
L.J. Natvik
January to May 1997

19. It’s not the end of the road…
Conclusion
It’s not the end of the road…