How to loop over years with libIGCM

Let's say that you want to run a 200Y long simulation, but you only want to use a few years of forcing data (we won't worry about why you want to do this at the moment). This is pretty easy to do with libIGCM by slightly modifying Josefine's instructions for running with analytical spinup, which I've mostly copied here.

NOTE: You have to loop over multiples of four years. If you do not, something strange happens in the leap years and only 11 out of the 12 months are printed in the history files sometimes. I talked to Josefine about it and we couldn't figure out why, so we're just using this as a fix. I have no idea how this works for things like FLUXNET validation.

NOTE2: Because of the previous note, take care to make sure your leap years line up in your looping years and your "real" years. I'll explain this more below.

NOTE3: You need libIGCM tag libIGCM_v2.0 (or trunk since revision 819).

How it works to use cyclic forcing

By adding the 2 variables CyclicBegin and CyclicEnd to config.card the cyclic fonctionality can be used. CyclicBegin is the first year of the forcing file to be used and CyclicEnd is the last year. So if I wanted to loop over years 1950-1953 (this is a multiple of four!), I would add the following to config.card:

CyclicBegin=1950
CyclicEnd=1953

libIGCM will use these variables to calculate the 2 new variables CyclicYear and !CyclicYear_p1. CyclicYear is the year within the forcing period [CyclicBegin, CyclicEnd]. !CyclicYear_p1 is the next coming cyclic year in the same interval (to be used for the PFTmap). These 2 new variables have to be used in orchidee_ol.card and sechiba.card to copy the correct forcing files.

Now let's say that I want to run for 10 total years (notice this is not a multiple of the number of cyclic years. If you are doing analytical spinup, it has to be, but not if you're just doing cyclic forcing. Again, we won't ask why you want to do this). We need to change config.card.

DateBegin=1950-01-01   # Same as CyclicBegin
DateEnd=1959-12-31     
PeriodLength=1Y

Having DateBegin? and CyclicBegin? be the same guarantees that the leap years will line up.

If we leave everything else the same, libIGCM isn't going to know to use the cyclic forcing files and it will just use the ones corresponding to the current date. This is because we only defined the variables in config.card; we haven't actually told libIGCM to use the variables anywhere yet. If we make the following modifications, then we are telling libIGCM exactly where to use the new variables.

COMP/orchidee_ol.card

[BoundaryFiles]
List=	(${R_BC}/OOL/${config_UserChoices_TagName}/NCC/ncc_for_${CyclicYear}.nc, forcing_file.nc)

COMP/sechiba.card

The PFTmap file must be copied using the variable !CyclicYear_p1 which is the next coming forcing year. In sechiba.card, remove PFTmap from section [InitialStateFiles]. Notice that you have to remove ",\" in the end of the line with floodplains.nc. This section becomes :

[InitialStateFiles]
List=       (${R_INIT}/SRF/${config_UserChoices_TagName}/routing.nc, .),     \
            (${R_INIT}/SRF/${config_UserChoices_TagName}/soils_param.nc, .), \
            (${R_INIT}/SRF/${config_UserChoices_TagName}/cartepente2d_15min.nc, .), \
            (${R_INIT}/SRF/${config_UserChoices_TagName}/floodplains.nc, .)

In section [BoundaryFiles] change following :

[BoundaryFiles]
#List=  (${R_BC}/SRF/${config_UserChoices_TagName}/PFTmap_1850to2005_AR5_LUHa.rc2/PFTmap_IPCC_${year_p1}.nc, .)
List=   ()
ListNonDel= ()

into

[BoundaryFiles]
List=  (${R_BC}/SRF/${config_UserChoices_TagName}/PFTmap_1850to2005_AR5_LUHa.rc2/PFTmap_IPCC_${CyclicYear_p1}.nc, PFTmap.nc)
ListNonDel= ()

Notice that you need to remove the additional "List= ()" line. No specific treatment in sechiba.driver will now be done for the PFTmap.nc file. It is not necessary to change in sechiba.driver. For the first year without restart file the PFTmap the will be used corrsponds to the next coming year. For all next coming years the correct PFTmap is used. It is not possible to come around this probleme while running with a period length one year.

PARAM/run.def ==

DRIVER_reset_time=y