1. Assimilation of Aqua Ocean Chlorophyll Data in a Global Three-Dimensional Model Watson Gregg NASA/Global Modeling and Assimilation Office

2. Motivations for Assimilation 1. Data use maximization 2. Parameter Estimation (model improvement) 3. State and Flux Estimation 4. Prediction

3. Atmospheric Forcing Data Radiative Model Layer Depths Circulation Model Biogeochemical Model Advection/ Diffusion Winds, SST Winds, ozone, rel. humidity, pressure, precip. H 2 O, cloud %, LWP, droplet radius, aerosols Layer Depths Current Velocities Particles Abundances Spectral Irradiance Temp. Spectral Radiance Primary Production Chlorophyll, Nutrients, Dust (Fe) Sea Ice NASA Ocean Biogeochemical Model (NOBM) Heat

4. Biogeochemical Model Diatoms Chloro- phytes Cyano- bacteria Cocco- lithophores Si NO 3 NH 4 Herbivores N/C Detritus Fe Silica Detritus Phytoplankton Nutrients Iron Detritus

5. Spectral Absorption Spectral Scattering m -1 ; m 2 mg -1 Wavelength (nm)

6. North Pacific North IndianEquatorial Indian North AtlanticNorth Central PacificNorth Central Atlantic Equatorial PacificEquatorial Atlantic South IndianSouth PacificSouth AtlanticAntarctic Day of Year Chlorophyll (mg m -3 ) Statistically positively correlated (P < 0.05) all 12 basins Gregg, W.W., 2002. Tracking the SeaWiFS record with a coupled physical/biogeochemical/radiative model of the global oceans. Deep-Sea Research II 49: 81-105. Gregg, W.W., P. Ginoux, P.S. Schopf, and N.W. Casey, 2003. Phytoplankton and Iron: Validation of a global three-dimensional ocean biogeochemical model. Deep-Sea Research II, 50: 3143-3169.

7. Assimilation of Satellite Ocean Chlorophyll Conditional Relaxation Analysis Method Advantages: Very strongly weighted toward data, less susceptible to model errors Fast Disadvantages Very susceptible to data errors 2 M =M,S 2

8. To keep assimilation model bounded requires: 1)Smoothing of data (25% monthly mean, 75% daily weight) 2) Increase model weighting relative to data 0.25 0.5 0.75 0.85 Model Weight (fraction)

9. M

10. Motivations for Assimilation 1. Data use maximization 2. Parameter Estimation (model improvement) 3. State and Flux Estimation 4. Prediction

11. Atmospheric Forcing Data Radiative Model Layer Depths Circulation Model Biogeochemical Model Advection/ Diffusion Winds, SST Winds, ozone, rel. humidity, pressure, precip. H 2 O, cloud %, LWP, droplet radius, aerosols Layer Depths Current Velocities Particles Abundances Spectral Irradiance Temp. Spectral Radiance Primary Production Chlorophyll, Nutrients, POC?, PIC? Dust (Fe) Sea Ice NASA Ocean Biogeochemical EOS Assimilation Model (OBEAM) Heat Red = EOS Data product Green = assimilated variable

12. Feb. 1, 2003

13. Motivations for Assimilation 1. Data use maximization 2. Parameter Estimation (model improvement) 3. State and Flux Estimation 4. Prediction

16. Annual RMS Log Error RMS mon = log 10 C assim – log 10 C aqua n √ RMS ann = ∑ ∑ RMS mon 12 X 100

19. North Pacific North IndianEquatorial Indian North AtlanticNorth Central PacificNorth Central Atlantic Equatorial PacificEquatorial Atlantic South IndianSouth PacificSouth AtlanticAntarctic Chlorophyll (mg m -3 ) Red = model monthly mean Diamonds = SeaWiFS monthly mean

20. Percent of Total Equatorial Pacific Diatoms Cocco Cyano Chloro 1997 200019981999 2001 20022003

22. Motivations for Assimilation 1. Data use maximization 2. Parameter Estimation (model improvement) 3. State and Flux Estimation 4. Prediction

24. Summary and Plans Initial assimilation results promising Need further analysis new methodologies Awaiting new SeaWiFS data Proceed on incorporation of MODIS/GMAO products

25. Atmospheric Forcing Data Radiative Model Layer Depths Circulation Model Biogeochemical Model Advection/ Diffusion Winds, SST Winds, ozone, rel. humidity, pressure, precip. H 2 O, cloud %, LWP, droplet radius, aerosols Layer Depths Current Velocities Particles Abundances Spectral Irradiance Temp. Spectral Radiance Primary Production Chlorophyll, Nutrients, POC?, PIC? Dust (Fe) Sea Ice NASA Ocean Biogeochemical EOS Assimilation Model (OBEAM) Heat Red = EOS Data product Green = assimilated variable