[MITgcm-devel] ptracers extension (new thread following "status")

Tue Jun 7 01:58:07 EDT 2011

Dear pkg/seaice and pkg/ptracers amigos,

the check-in I did last night, within pkg/seaice, is meant as another step towards a generic seaice 
tracers pkg (or a new component to pkg/ptracers; see earlier devel emails). As compared with 
the current pkg/seaice codes that handle HSALT and IceAgeTr, the main differences are 
1) the intensive (i.e. 'not effective') variables approach, in analogy with ocn/atm/ptracers components.
2) the detailed accounting for the sequence of seaice growth/melt/flooding processes (d_HEFF*).
I have good hope that you will find the new code to be sensible. 

The intention is, in a reasonable time frame, to replace the old codes and, potentially,
later, to move the new codes in a different pkg where it could also be used by pkg/thsice. 
This being said, there is still a fair amount of work needed before any of this could 
happen, so dont get too nervous ;-) Also the choice to proceed by addition (of 
ALLOW_SITRACER to the existing SEAICE_VARIABLE_SALINITY & SEAICE_AGE) 
rather than direct substitution is obviously intended to facilitate the transition.

Amongst those things that still need work, sooner or later: area tracers, pickup/initial conditions,
run time parameters, diags, TAF, advection variants, ptracers and thsice connections.

Cheers,
Gael

Presently, in case you wanted to give the new code a try, do the following
in SEAICE_OPTIONS.h add
	#define ALLOW_SITRACER
	#define ALLOW_SITRACER_ADVCAP
	#define ALLOW_SITRACER_DIAG
in data.seaice add
	&SEAICE_PARM03
	siTrName(1)           = 'age',
	siTrName(2)           = 'salinity',
	siTrName(3)           = 'one',
in data.diagnostics add something like
	fields(1,4)  = 'UDIAG2  ',
	fileName(4) = 'sitrAdvDiag',
	frequency(4) = 2592000.,
	fields(1,5)  = 'UDIAG1  ',
	fileName(5) = 'sitrPhyDiag',
	frequency(5) = 2592000.,

On Feb 25, 2011, at 10:59 AM, Jean-Michel Campin wrote:

> Hi,
> 
> I thought it would be more clear to start a new thread, since
> this one ("Status") is still going on with pkg/seaice discussion.
> 
> I support Dimitris' plan, because right now ptracers only 
> handle 3-D tracers which are advected by uVel,vVel,wVel.
> To have an extension to seaice tracer require a new set of
> arrays + S/R  and it's not completely clear how general
> it should be (we may want to enable a 2-D passive tracer
> to be advected with "surface current", also with cheapAML winds, 
> or with thsice related variables, but this requires some options
> regarding the flow field to use and which mask/volume to use).
> 
> The advection S/R (pkg/generic) can be used for any tracer /
> flow field (or at least, can/should be changed if it's not fully 
> the case), and it's already a good point. 
> 
> And it makes sense for me to have a 1rst try within pkg/seaice first,
> then to see how general this 2-D passive tracer case is (flow field,
> mask/volume) and then to move it to pkg/ptracers when things are 
> more clear regarding extension to other 2-D passive-tracer cases.
> 
> Cheers,
> Jean-Michel
> 
> On Thu, Feb 24, 2011 at 10:29:06AM -0500, Chris Hill wrote:
>> Hi All,
>> 
>> For what its worth, I think we should be getting this stuff going
>> within a ptracers framework.
>> It would be a shame not to do that.
>> 
>> Chris
>> 
>> On Thu, Feb 24, 2011 at 9:42 AM, Menemenlis, Dimitris (3248)
>> <Dimitris.Menemenlis at jpl.nasa.gov> wrote:
>>> Gael, thank you for message. I know that your two scenarios were the starting point of discussions for ice age tracer but I don't yet know what Pierre decided to implement as I did not take part in discussions about this yesterday.
>>> 
>>> Could you remind me what "extensive" and "intensive" mean. Do they mean area-based as opposed to volume-based? Eventually we should implement both. The first is more directly observable while the second seems to be a more meaningful measure of age.
>>> 
>>> A third useful age diagnostic, down the road, will be to use several simultaneous age tracers that categorize fraction of first year ice, second year ice, etc., in each cell.
>>> 
>>> The eventual hope is that we will be able to use one or more of the age diagnostics to better parameterize sea ice properties like drag, thickness distribution, and ice strength.
>>> 
>>> Dimitris Menemenlis
>>> 818-625-6498
>>> 
>>> On Feb 24, 2011, at 1:01 AM, "Gael Forget" <gforget at MIT.EDU> wrote:
>>> 
>>>> Hi Dimitris and Patrick,
>>>> 
>>>> a few thoughts regarding ice tracers (salinity and age for now).
>>>> 
>>>> My recollection is that there are currently two issues regarding the age tracer:
>>>> (1) it is advected as extensive, whereas it is otherwise treated as intensive.
>>>> (2) the lack of detail regarding the effect of individual growth/melt terms on age.
>>>> 
>>>> With regard to generic ice tracers a choice will have to be made between
>>>> extensive (as in SEAICE_SALINITY) and intensive (as in SEAICE_AGE).
>>>> I wrote down the two scenarios equations for Pierre so he knows how to
>>>> proceed either way for the age tracer. I dont have a strong preference but
>>>> 'intensive' would imply  a division by HEFF during advection (as opposed
>>>> to diagnostics) that could prove problematic. So I am wondering whether
>>>> it would it be good time to switch SEAICE_AGE to extensive.
>>>> 
>>>> Cheers,
>>>> Gael
>>> 
>>> _______________________________________________
>>> MITgcm-devel mailing list
>>> MITgcm-devel at mitgcm.org
>>> http://mitgcm.org/mailman/listinfo/mitgcm-devel
>>> 
>> 
>> _______________________________________________
>> MITgcm-devel mailing list
>> MITgcm-devel at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-devel
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel