HYCOM
423
Comparison with drifting buoys: These tests will evaluate the models’
ability to produce ocean currents that yield drifter and ARGO floats
trajectories similar to observations.
Mixed Layer Depth (MLD) (analysis, forecast, simulation without ocean
data assimilation): Model analyses, forecasts, and simulations will be
compared to mixed layer depths from profile data (e.g. XBTs, ARGO floats,
CTDs, and moored buoys) and to an MLD climatology.
Event comparisons: Independent data are used for qualitative and
quantitative evaluation of prediction system skill in nowcasting and
forecasting specific oceanic events and features. A classical example is the
impact of hurricanes on the ocean circulation (Zamudio et al., 2002).
Comparisons of surface height and temperature with ocean color imagery
can at times provide clear and dramatic qualitative model assessment
(Chassignet et al, 2005).
The near real-time North Atlantic basin model outputs are made
available to the community at large within 24 hours via the Miami Live
Access Server (LAS) (http://hycom.rsmas.miami.edu/las). Specifically, the
LAS supports model-data and model-model comparisons; provides HYCOM
subsets to coastal or regional nowcast/forecast partners as boundary
conditions, and increases the usability of HYCOM results by application
providers .
4.
Outlook
The long term goal is an eddy-resolving, fully global ocean prediction
system with data assimilation based on HYCOM to be transitioned to the
Naval Oceanographic Office at 1/12q equatorial (~7 km mid-latitude)
resolution in 2007 and 1/25q resolution by 2011. This paper summarizes the
present status of the HYCOM effort and illustrates its capabilities. The
present systems are a first step towards the fully global 1/12° HYCOM
prediction system. The size of the problem makes it very difficult to use
sophisticated assimilation techniques. Some of these methods can increase
the cost of running the model by a factor of 100. It is, however, important to
evaluate the performance of these advanced data assimilation techniques.
Several additional techniques for assimilating data into HYCOM are already
in place or are in the process of being implemented. These techniques vary
in sophistication and computational requirements and include: NRL Coupled
Ocean Data Assimilation (NCODA), Singular Evolutive Extended Kalman
(SEEK) filter, Reduced Order Information Filter (ROIF), Ensemble Kalman
Filter (EnKF), Reduced Order Adaptive Filter (ROAF) (including adjoint),
and the 4D-VAR Representer method.
“
”
423
Comparison with drifting buoys: These tests will evaluate the models’
ability to produce ocean currents that yield drifter and ARGO floats
trajectories similar to observations.
Mixed Layer Depth (MLD) (analysis, forecast, simulation without ocean
data assimilation): Model analyses, forecasts, and simulations will be
compared to mixed layer depths from profile data (e.g. XBTs, ARGO floats,
CTDs, and moored buoys) and to an MLD climatology.
Event comparisons: Independent data are used for qualitative and
quantitative evaluation of prediction system skill in nowcasting and
forecasting specific oceanic events and features. A classical example is the
impact of hurricanes on the ocean circulation (Zamudio et al., 2002).
Comparisons of surface height and temperature with ocean color imagery
can at times provide clear and dramatic qualitative model assessment
(Chassignet et al, 2005).
The near real-time North Atlantic basin model outputs are made
available to the community at large within 24 hours via the Miami Live
Access Server (LAS) (http://hycom.rsmas.miami.edu/las). Specifically, the
LAS supports model-data and model-model comparisons; provides HYCOM
subsets to coastal or regional nowcast/forecast partners as boundary
conditions, and increases the usability of HYCOM results by application
providers .
4.
Outlook
The long term goal is an eddy-resolving, fully global ocean prediction
system with data assimilation based on HYCOM to be transitioned to the
Naval Oceanographic Office at 1/12q equatorial (~7 km mid-latitude)
resolution in 2007 and 1/25q resolution by 2011. This paper summarizes the
present status of the HYCOM effort and illustrates its capabilities. The
present systems are a first step towards the fully global 1/12° HYCOM
prediction system. The size of the problem makes it very difficult to use
sophisticated assimilation techniques. Some of these methods can increase
the cost of running the model by a factor of 100. It is, however, important to
evaluate the performance of these advanced data assimilation techniques.
Several additional techniques for assimilating data into HYCOM are already
in place or are in the process of being implemented. These techniques vary
in sophistication and computational requirements and include: NRL Coupled
Ocean Data Assimilation (NCODA), Singular Evolutive Extended Kalman
(SEEK) filter, Reduced Order Information Filter (ROIF), Ensemble Kalman
Filter (EnKF), Reduced Order Adaptive Filter (ROAF) (including adjoint),
and the 4D-VAR Representer method.
“
”