Dear Michael et al.

The slaboc option is quite straightforward to use .

1) Build a climatology of q-fluxes ( compensating for the deviation of the SST calculated using a simple mixed layer energy budget , and the observations you force normally the model with e.g OISST) :
you do a first simulation using

islab_ocean=1
in physicsparam
and
&slabocparam
do_restore_sst = .true.
do_qflux_adj = .false.

This will produce a monthly q-flux file in your input directory call domname_SOM.YYYYMMDDHH.nc. You need at least to run one year to have a complete annual cycle, but it is advised to run for several year to build a proper climatology. In this run the slaboc SST are restored to observations , so it should be close to run with purely forces SST. If you check it you should see normally the bigger value of q-fluxes ( +or -)where the slab-ocean prognostic SST are diverging strongly from reality, i.e. where the ocean dynamic plays a lot in determining SST 's, e.g upwelling regions.

Then you can proceed to "adjusted runs" runs where you will use the SST calculated from the slab ocean energy budget accounting fro the corrective energy fluxes saved in phase one ( the q flux basically drives the seasonal cycle insuring your SST stays realistic, while allowing shorter time intearctions with the atmosphere) ..

islab_ocean=1
in physicsparam
and
&slabocparam
do_restore_sst = .false.
do_qflux_adj = .true.

Typically you run a control "adjusted" run, and then some experiment "adjusted" runs where you modify e.g. landuse, aerosol, convection schemes, etc .

Overall this is simple to operate ,the inconvenient is that you need to perform one more run to produces a reference q-flux climatology ..it has also been tested mostly over the Indian domain .

Good to know, there are some parameters that might be sensitive ( but never really tested ) in the slaboc . e.g. the time scale for restoring ( fixed to five days ) , or the depth of the mixed layer ( fixed to 50m) . We have a monthly climatology of global mixed layer depth that would be good to use, but never had time to integrate this ...

Fabien

On 11/11/15 16:40, Michael Notaro wrote: