Modeling Ocean Circulation
39
Figure 3.6. Claes Rooth of the University of Miami was a visitor in Princeton in 1981. His visit inspired
work on isopycnal mixing and multi-equilibrium solutions in ocean models. Photo taken 2005.
surface. Manabe et al. (1991) use a stopgap remedy called the “flux adjusted” coupled
model. In this parametrization the flux of heat and fresh water at the air–sea interface
consists of two parts: one part is fixed at the rate which is obtained by an ocean-only
model with realistic boundary conditions, and a second part is variable, depending
on the interactive sea surface temperature and air temperature in the atmosphere just
above the surface. While this remedy reduces the generality of the coupled model, it
allowed the coupled model to have a realistic response to perturbations such as the
effect of changes in greenhouse gases. Years of struggle finally paid off. At the time
of the first IPCC (International Panel on Climate Change) report GFDL was the only
laboratory with a global coupled ocean–atmosphere model.
Claes Rooth and Juergen Willebrand also contributed to the very important idea
for improving the lateral mixing parametrization in the ocean circulation model. It
had been pointed out that lateral mixing in the real ocean is largely along isopycnal surfaces, while in our GFDL model parametrization mixing largely took place
on horizontal surfaces. The result was a spurious mixing across tilted density surfaces, unsupported by field data. As far back as the National Academy meeting in
New Hampshire in 1972 (NAS,1975) George Veronis pointed out that this spurious
39
Figure 3.6. Claes Rooth of the University of Miami was a visitor in Princeton in 1981. His visit inspired
work on isopycnal mixing and multi-equilibrium solutions in ocean models. Photo taken 2005.
surface. Manabe et al. (1991) use a stopgap remedy called the “flux adjusted” coupled
model. In this parametrization the flux of heat and fresh water at the air–sea interface
consists of two parts: one part is fixed at the rate which is obtained by an ocean-only
model with realistic boundary conditions, and a second part is variable, depending
on the interactive sea surface temperature and air temperature in the atmosphere just
above the surface. While this remedy reduces the generality of the coupled model, it
allowed the coupled model to have a realistic response to perturbations such as the
effect of changes in greenhouse gases. Years of struggle finally paid off. At the time
of the first IPCC (International Panel on Climate Change) report GFDL was the only
laboratory with a global coupled ocean–atmosphere model.
Claes Rooth and Juergen Willebrand also contributed to the very important idea
for improving the lateral mixing parametrization in the ocean circulation model. It
had been pointed out that lateral mixing in the real ocean is largely along isopycnal surfaces, while in our GFDL model parametrization mixing largely took place
on horizontal surfaces. The result was a spurious mixing across tilted density surfaces, unsupported by field data. As far back as the National Academy meeting in
New Hampshire in 1972 (NAS,1975) George Veronis pointed out that this spurious