Ocean Modeling

Research focused upon improved model representation of ocean processes and particularly the processes governing sea surface temperature, upper ocean heat content, and salinity variability including air-sea exchanges, heat-flux, lateral ocean advection, and entrainment at the base of the ocean mixed layer that play a significant role in controlling short-term variability in ocean and coastal circulations as well as long-term variations. It also includes modeling of the ocean from the surface to the ocean floor to improve understanding and, eventually, forecasting of climate variability and climate change.

Below are two of the many projects that are classified on this theme:

  1. Assessing the Sensitivity of Northward Heat Transport/Atlantic Meridional Overturning Circulation to Forcing in Existing Numerical Model Simulations by S. Dong (UM/CIMAS); M. Baringer, G. Goni and G. Halliwell (NOAA/AOML)

  2. Observing System Simulation Experiments (OSSEs) in the Gulf of Mexico by V. Kourafalou (UM/RSMAS); M. Le Hénaff and P. Ortner (UM/CIMAS); R. Atlas and G. Halliwell (NOAA/AOML); for more details in OSSEs pls. see "Ocean Modeling and OSSE Center (OMOC)"

Representative Projects

Assessing the Sensitivity of Northward Heat Transport/Atlantic Meridional Overturning Circulation to Forcing in Existing Numerical Model Simulations

S. Dong (UM/CIMAS); M. Baringer, G. Goni and G. Halliwell (NOAA/AOML)

Long Term Research Objectives and Strategy to Achieve Them:

Objectives: To investigate the mechanisms underlying the observed differences in the role of Ekman and geostrophic transports in the Atlantic Meridional Overturning Circulation (AMOC) and the net northward heat transport in both the North and South Atlantic on seasonal to longer time scales, and to diagnose the causes for the inconsistency between their observed variability and that demonstrated in the numerical model simulations.

Strategy: Combine data analyses and numerical model outputs.


The Atlantic Meridional Overturning Circulation (AMOC) is characterized by a northward flow of warm water in the upper layers from the tropics and the South Atlantic into the North Atlantic, and a southward return flow of cold water at depth. This large-scale ocean circulation is the main route for
the global ocean heat conveyor belt circulation in the Atlantic, and it is an important benchmark for Earth’s climate. The AMOC carries 25% of the total global ocean-atmosphere northward heat flux. The majority of this heat is lost to the atmosphere in the mid-latitudes where warm water meets cold, dry continental air masses. The AMOC regulates and maintains the meridional temperature distribution in the Atlantic. Variations in the AMOC and its associated ocean heat transport are coupled to atmospheric heat transport variations through the mechanism of Bjerknes compensation. As a result, changes in the AMOC can have a direct and pronounced impact on variety of climate phenomena, such as African and Indian monsoon rainfall, hurricane activity, climate over the North America and Western Europe. Thus, understanding and monitoring AMOC variability are crucial for improving our knowledge of the mechanisms of climate system and for assessing future climate change.

Figure 1. (a) Time series of the AMOC (black) and contributions from the geostrophic (red) and Ekman (green) components. (b) West-to-East cumulative transports (in Sv) averaged over the 17 AX18 transects. The dashed lines indicate the separation points for the boundary currents and interior. (c) Contributions of the western (red) and eastern (green) boundaries and interior (blue) to the AMOC. (d) The total cumulative volume transport (black) from the surface to ocean bottom, and those in the western (red) and eastern (green) boundaries and interior (blue) for the December 2004 transect.


Figure 2. Scatter plots of the (a) AMOC, and its (b) geostrophic and (c) Ekman contributions versus month. (d) Contributions of the western (circles) and eastern (triangles) boundaries and the interior (stars) to the AMOC. The black lines show the annual cycle. Monthly climatological (dashed) wind stress curl (e) and zonal wind stress (f) averaged in the subtropical south Atlantic (box indicated in (d)). Solid lines in (e) and (f) are the corresponding annual harmonic.

Variability of the AMOC and its effect on the net northward heat transport in the South Atlantic were investigated using a trans-basin expendable bathythermograph (XBT) high-density line at 35S (AX18). The mean time-mean AMOC was 17.92.2 Sv during 2002-2007. Although geostrophic transport dominates the time-mean AMOC, both geostrophic and Ekman transports are important in explaining the AMOC variability. The contributions of geostrophic and Ekman transports to the AMOC both show annual cycles, but they are out of phase, resulting in weak seasonal variability of the AMOC. Northward heat transport variability is significantly correlated with the AMOC, in that a 1Sv increase in the AMOC would yield a 0.050.01 PW increase in the northward heat transport. Partitioning transport at the western and eastern boundaries suggests that, to quantify changes in the AMOC and total northward heat transport, it is essential to monitor all three regions.

Figure 3. Time series of (a) the NHT (red, right axis) and the AMOC (black, left axis), and (b) the NHT (red) and contributions from the barotropic (blue), overturning (black), and horizontal (green) components. (c) Scatter plots of the overturning (black) and horizontal (green) heat fluxes versus month. Units are PW for heat transports and Sv for volume transports.


Observing System Simulation Experiments (OSSEs) in the Gulf of Mexico

V. Kourafalou (UM/RSMAS); M. Le Hénaff and P. Ortner (UM/CIMAS); R. Atlas and G. Halliwell (NOAA/AOML)

for more details in OSSEs pls. see "Ocean Modeling and OSSE Center (OMOC"


Long Term Research Objectives and Strategy to Achieve Them:

Objectives : This project facilitated the initial development of Observing System Simulation Experiments (OSSEs), which are now a main activity under the Joint NOAA-AOML and UM-RSMAS Ocean Modeling and OSSE Center (OMOC). OSSEs are a tool to evaluate the impact of specific observation systems on our ability to accurately hindcast and forecast important physical processes prior to actually collecting observations. Since observing systems are expensive to deploy and maintain, a-priori understanding of the impact of different observing strategies is a crucial component of NOAA’s Integrated Ocean Observing Systems initiative. The goal of the OSSE initial development in the Gulf of Mexico (GoM) is to assess the expected performances of various network systems to monitor the Loop Current (LC) dynamics.

Strategy: The development of high resolution, data evaluated models that can perform reliable simulations of the GoM mesoscale variability and be integrated with observations toward data assimilative forecasts.


This project has focused on the development of a regional ocean OSSE methodology, in collaboration with AOML (R. Atlas and G. Halliwell), a collaboration that evolved in the OSSE activities within the OMOC. In particular, we advanced a “toolbox” that can be shared by the RSMAS and AOML ocean OSSE research groups and we have been working toward a nature run in the Gulf of Mexico with suitable data evaluation and performance metrics. In addition, this project is connected to an international initiative on nested, data assimilative coastal and regional models, namely the Global Ocean Data Assimilation Experiment (GODAE) through the Coastal and Shelf Seas Task Team activities of the GODAE/OceanView.

The LC is the main dynamical feature in the GoM. It is highly variable in time, from a retracted position where it flows directly from the Yucatan Strait to the Straits of Florida, to an extended position where it flows far north into the GoM before turning southeastward towards the Straits of Florida. When extended, the LC eventually sheds a large, warm-core anticyclonic eddy or ring (Loop Current Eddy, LCE) that drifts westward in the GoM. The LC then returns to its retracted position. This variability appears to be highly determined by the LC frontal dynamics. Small cyclonic eddies at the edge of the LC (Fig. 1) play an active role in the variability of the LC and in the LCE shedding. This role has been highlighted during the recent Deepwater Horizon oil spill, when such a small cyclonic eddy advected pollutants southward, before they were trapped within LCE Franklin that formed at that time. OSSEs need to address a specific question that the monitoring and prediction systems will address. We are interested in better understanding, representing and predicting the interactions between the LC and the frontal eddies, especially during the LCE shedding process.

Figure 1. Daily model Sea Surface Height (SSH) on (a) December 18, 2005, (b) January 2, 2006, depicting the synergy between the cyclonic eddies (low SSH, blue colors) and the formation/separation of an anticyclonic Loop Current Eddy (hiugh SSH, red colors). Black dotted lines indicate 200, 2000 and 3000 m isobaths.

To address this issue, we have set up a high-resolution model of the circulation of the Gulf of Mexico (GoM-HYCOM). This model has been run to simulate five years of this circulation, giving us insights on the full 3D dynamics. This has allowed us to better characterize the interactions between the LC and its associated frontal eddies, which lead to the LCE separation from the LC. In addition, an ensemble of 40 simulations with different forcing conditions in the boundary inflow has been run during an LCE shedding episode. This ensemble of simulations has allowed us to understand our model uncertainty regime, associated to the growth of frontal eddies and their migration. Observations from surface drifters in the LC have confirmed some of our results.

In parallel, an advanced data assimilation (DA) scheme has been set up for use with our ocean model. It will be used to integrate simulated observations within the model; this will allow us to assess the expected performance of hypothetical observation networks, thus helping in array design. The
networks to be tested will be dedicated to the observation of the LC frontal eddy dynamics. After the recent Deepwater Horizon oil spill event, we have also performed numerical experiments to model the displacement of water particles initiated at the observed oil spill patch. These simulations are based on the Lagrangian drift module from a tool originally developed to study fish larvae dispersion. We are now working on improving the modeling of the oil within this tool, to better describe the distribution of oil and gas at depth.