[MITgcm-support] "how to proper use both net heat flux and sstrelaxation?? "

[Martin Losch]

Hi Li Zhi,

the relaxation terms are used to “pull” the model towards “climatology”, i.e. something that is, in the broadest of all senses, “observed”. There’s a lot of literation about this, because it also affects the type of boundary conditions (mixed bcs or not).

Usually, one would use weak restoring, with a time scale of months. In the other extreme: for climsss/sstTauRelax = deltaT you practically replace the surface field with the “climsst/sssField”

Another thing about relaxation: when you relax to a time varying field, e.g. monthly climatology, with a long time scale (months) you introduce a phase shift (in the same way as in a forced oscillator), so I tend to relax towards constant (in time) fields, if at all.

Martin

On May 27, 2014, at 5:05 PM, 李志远 <oceanlizy at 163.com> wrote:

>  But,I am still some pazzle that how to control climsstTauRelax and climsssTauRelax ? If I use small climsstTauRelax ,such as 2days or more , will the surface  temperature of model output theta be close to the given sst ? 
>