A fully coupled physical and biogeochemical ocean data assimilation system is tested in a realistic configuration of the California Current System using the Regional Ocean Modeling System. In situ measurements for sea surface temperature and salinity as well as satellite observations for temperature, sea level and chlorophyll are used for the year 2000. Initial conditions of the combined physical and biogeochemical state are adjusted at the start of each 3-day assimilation cycle. Data assimilation results in substantial reduction of root-mean-square error (RMSE) over unconstrained model output. RMSE for physical variables is slightly lower when assimilating only physical variables than when assimilating both physical variables and surface chlorophyll. Surface chlorophyll RMSE is lowest when assimilating both physical variables and surface chlorophyll. Estimates of subsurface, nitrate and chlorophyll show modest improvements over the unconstrained model run relative to independent, unassimilated in situ data. Assimilation adjustments to the biogeochemical initial conditions are investigated within different regions of the California Current System. The incremental, lognormal 4-dimensional data assimilation method tested here represents a viable approach to coupled physical biogeochemical state estimation at practical computational cost.
Bibliographical noteFunding Information:
We are grateful for support from the Gordon and Betty Moore Foundation and from the National Oceanographic and Atmospheric Administration (NA10OAR4320156) that enabled this research. SST data was obtained from http://las.pfeg.noaa.gov/oceanWatch/oceanwatch.php and surface chlorophyll data from http://oceandata.sci.gsfc.nasa.gov/SeaWiFS/Mapped/Daily/9km/ . Three anonymous reviewers significantly improved the manuscript with several helpful comments.
© 2016 Elsevier Ltd
All Science Journal Classification (ASJC) codes
- Computer Science (miscellaneous)
- Geotechnical Engineering and Engineering Geology
- Atmospheric Science