
The range of stability of the LeapFrog scheme can be extended by a factor of two by introducing a semiimplicit computation of the hydrostatic pressure gradient term [Brown and Campana, 1978]. Instead of evaluating the pressure at , a linear combination of values at , and is used (see § 6.4.5). This technique, controlled by the nn_dynhpg_rst namelist parameter, does not introduce a significant additional computational cost when tracers and thus density is time stepped before the dynamics. This time step ordering is used in NEMO (Fig.3.1).
This technique, used in several GCMs (NEMO, POP or MOM for instance), makes the LeapFrog scheme as efficient ^{3.1} as the ForwardBackward scheme used in MOM [Griffies et al., 2005] and more efficient than the LFAM3 scheme (leapfrog time stepping combined with a third order AdamsMoulthon interpolation for the predictor phase) used in ROMS [Shchepetkin and McWilliams, 2005].
In fact, this technique is efficient when the physical phenomenon that limits the timestep is internal gravity waves (IGWs). Indeed, it is equivalent to applying a time filter to the pressure gradient to eliminate high frequency IGWs. Obviously, the doubling of the timestep is achievable only if no other factors control the timestep, such as the stability limits associated with advection, diffusion or Coriolis terms. For example, it is inefficient in low resolution global ocean configurations, since inertial oscillations in the vicinity of the North Pole are the limiting factor for the time step. It is also often inefficient in very high resolution configurations where strong currents and small grid cells exert the strongest constraint on the time step.
Gurvan Madec and the NEMO Team
NEMO European Consortium20170217