(back up to Science Work Robin)
Water/Ocean Tracer conservation
(basically copied from the old muddle)
Rigid lid ocean models use virtual salt fluxes at the surface to parameterise water fluxes, as they cannot change their volume. The salt fluxes are calculated by multiplying the water flux applied by a salinity value. Assuming global water fluxes are conserved, the virtual salt fluxes can only be guaranteed to conserve if a globally constant reference salinity is used in the calculation - this is non-ideal, since local salinity changes are proportional to the salinity of the water already there (adding freshwater to freshwater should not produce a salinity change, whereas adding freshwater to salty should). Nonetheless the reference salinity was used in HadCM3, with consequent distortion of the virtual salt fluxes, but this did not have serious effects on the simulation. However, switching on the reference salinity in FAMOUS led to rapid weakening of the Atlantic MOC. Using the local salinity without any correction is also unacceptable, because the consequent large salinity drift would be serious over the long periods of time for which FAMOUS may be run.
The correction applied in FAMOUS is calculated thus: for open ocean points, the water fluxes and virtual salt fluxes are accumulated every timestep. Once a year these are summed and a globally constant correction is derived. This is then applied over the course of the next year so that the long term salinity drift in the ocean matches any long term drift in the water fluxes applied (drift in water fluxes may come about from the accumulation of snow on land, for example). The correction is applied to every ocean box, not just at the surface: adtan has already run long enough that the current drift is fairly uniform with depth, and applying the correction only at the surface would result in a strong freshening of surface salinity (whilst the deep ocean still gets saltier) that impacts the MOC strength.