424
ERIC P. CHASSIGNET ET AL.
NCODA is an oceanographic version of the multivariate optimum
interpolation (MVOI) technique widely used in operational atmospheric
forecasting systems. A description of the MVOI technique can be found in
Daley, (1991). The ocean analysis variables in NCODA are temperature,
salinity, geopotential (dynamic height), and velocity. The horizontal
correlations are multivariate in geopotential and velocity, thereby permitting
adjustments to the mass field to be correlated with adjustments to the flow
field. NCODA assimilates all available operational sources of ocean
observations. This includes along track satellite altimeter observations,
MCSST and in situ observations of SST and SSS, subsurface temperature
and salinity profiles from BT’s and profiling floats, and sea ice
concentration.
Both the SEEK filter (Pham et al , 1998) and ROIF (Chin et al., 1999) are
sequential in nature, implying that only past observations can influence the
current estimate of the ocean state and are especially well suited for large
dimensional problems. The ROIF assumes a tangent linear approximation to
the system dynamics, while the SEEK filter can use the non-linear model to
propagate the error statistics forward in time (Ballabrera et al., 2001).
Besides the NCODA, SEEK and ROIF methods, other techniques such as
the EnKF and the ROAF are also being evaluated. Because of their cost,
they are presently being evaluated mostly within coastal HYCOM
configurations or in specific limited areas of high interest. The NCODA and
SEEK techniques are being considered as the next generation data
assimilation to be used in the near real-time system.
Development of the global HYCOM prediction system is presently
taking place and includes model development, data assimilation, and ice
model embedment. The model configuration is fully global with the Los
Alamos CICE ice model embedded and will run at three resolutions: ~60
km, ~20 km and ~7 km at mid-latitudes with the NCODA data assimilation.
As stated above, some of the more expensive data assimilation techniques,
while impractical over a high resolution global domain, can be used in
subregions of the global model domain where there is special interest or
where they provide particular value added.
Acknowledgements
This work was sponsored by the National Ocean Partnership Program
(NOPP), the Office of Naval Research (ONR), and the Operational Effects
Programs (OEP) Program Office, PMW 150 through the following projects:
NOPP HYCOM Consortium for Data-Assimilative Ocean Modeling, NOPP
U.S. GODAE: Global Ocean Prediction with the HYbrid Coordinate Ocean
.
ERIC P. CHASSIGNET ET AL.
NCODA is an oceanographic version of the multivariate optimum
interpolation (MVOI) technique widely used in operational atmospheric
forecasting systems. A description of the MVOI technique can be found in
Daley, (1991). The ocean analysis variables in NCODA are temperature,
salinity, geopotential (dynamic height), and velocity. The horizontal
correlations are multivariate in geopotential and velocity, thereby permitting
adjustments to the mass field to be correlated with adjustments to the flow
field. NCODA assimilates all available operational sources of ocean
observations. This includes along track satellite altimeter observations,
MCSST and in situ observations of SST and SSS, subsurface temperature
and salinity profiles from BT’s and profiling floats, and sea ice
concentration.
Both the SEEK filter (Pham et al , 1998) and ROIF (Chin et al., 1999) are
sequential in nature, implying that only past observations can influence the
current estimate of the ocean state and are especially well suited for large
dimensional problems. The ROIF assumes a tangent linear approximation to
the system dynamics, while the SEEK filter can use the non-linear model to
propagate the error statistics forward in time (Ballabrera et al., 2001).
Besides the NCODA, SEEK and ROIF methods, other techniques such as
the EnKF and the ROAF are also being evaluated. Because of their cost,
they are presently being evaluated mostly within coastal HYCOM
configurations or in specific limited areas of high interest. The NCODA and
SEEK techniques are being considered as the next generation data
assimilation to be used in the near real-time system.
Development of the global HYCOM prediction system is presently
taking place and includes model development, data assimilation, and ice
model embedment. The model configuration is fully global with the Los
Alamos CICE ice model embedded and will run at three resolutions: ~60
km, ~20 km and ~7 km at mid-latitudes with the NCODA data assimilation.
As stated above, some of the more expensive data assimilation techniques,
while impractical over a high resolution global domain, can be used in
subregions of the global model domain where there is special interest or
where they provide particular value added.
Acknowledgements
This work was sponsored by the National Ocean Partnership Program
(NOPP), the Office of Naval Research (ONR), and the Operational Effects
Programs (OEP) Program Office, PMW 150 through the following projects:
NOPP HYCOM Consortium for Data-Assimilative Ocean Modeling, NOPP
U.S. GODAE: Global Ocean Prediction with the HYbrid Coordinate Ocean
.