High-Resolution
Global Coupled Ocean/Sea Ice Modeling
The objective of this project is to couple a high-resolution
ocean general circulation model with a high-resolution dynamic-thermodynamic
sea ice model in a global context. Currently, such simulations
are typically performed with a horizontal grid resolution of
about 1 degree. At this resolution (about 30 to 50 km in the
polar regions), the ocean model cannot resolve very narrow current
systems (including fronts and turbulent eddies) that play a
crucial role in the transport of heat and salt in the global
ocean. Similarly, lower-resolution sea ice models cannot resolve
important dynamics that occur in regions of complicated topography
(such as the Canadian Archipelago).
|
|
|
|
Figure 2
High-resolution (1/10 degree) POP ocean model currents
at 50m depth. Blue = 0; red > 150 cm/s. |
|
This project is running a global ocean circulation model
with horizontal resolution of approximately 1/10th degree—between
11 km and 2.5 km (Figure 2). This is the highest-resolution
simulation ever attempted with a such a realistic model. This
configuration has dimensions of 3600 x 2400 x 40, resulting
in 177 million active ocean grid points (some grid points
are on land). The code being used is the Parallel Ocean Program
(POP), developed at LANL under the Department of Energy’s
CHAMMP program. At NERSC, 448 processors are used to run the
model. One year can be simulated in about eight wall-clock
days (86,000 processor hours), generating over 500 GB of output.
Eight model years have been run to date, with a goal of 30–50
years. After the ocean simulation has run for 10–15
model years, it will be coupled with a sea ice model to more
accurately simulate the polar circulation.
The interaction of the ocean and overlying sea ice in global
coupled numerical models is poorly understood, though very
important. When ocean water freezes into sea ice, salt is
released into the upper ocean, making it more dense. Conversely,
when the ice melts, it creates a layer of fresh water that
is less dense than the underlying ocean. This delicate balance
between melting and freezing is very difficult to simulate
with coarse grids. In particular, high vertical resolution
is needed near the surface to simulate this salinity balance
correctly. High horizontal resolution is required to properly
simulate the current systems that advect these salinity anomalies
into the open ocean. Inaccuracies in the surface ocean properties
due to poor representation of ocean-ice interaction can have
wide-ranging global consequences. Most notable is the possibility
that too much fresh surface water can inhibit vertical convection
in the northern seas (since it is less dense than the salty
water beneath it), which then disrupts the entire global heat
budget. Coarse-resolution simulations have found that the
circulation and heat budget are extremely sensitive to the
way sea ice is prescribed in ocean-only runs. The best tool
for simulating the global circulation accurately is a high-resolution,
fully coupled ocean-sea ice model.
INVESTIGATORS
M. E. Maltrud and E. C. Hunke, Los Alamos National Laboratory;
J. L. McClean, Naval Postgraduate School.
PUBLICATIONS
In preparation.
URL
http://www.lanl.gov/orgs/t/t3/codes/pop.shtml
|