[MITgcm-support] rbcs advice

Tue Mar 11 10:26:26 EDT 2014

Hi Martin,

On Mar 11, 2014, at  0:59 AM, Martin Losch <Martin.Losch at awi.de> wrote:

> are you going to implement that for yourself? You probably have already. If so, you can send me the modified files, and I’ll check them in.

I tried, but I guess I don't understand how to use namelists because the change didn't do anything.  Hard coding the time offset into obcs_fields_load.F did work. I must admit my FORTRAN skills are not what they should be...

> In fact, the obcs package can use the exf-package, which already has all of this functionality. The exf-package is more complicated to use at first, but once you get your head around it, it’s much more flexible than the default way of reading forcing and obcs fields and should be the preferred way of reading forcing fields. When I think about it, rather than adding a new flag to obcs I should add the exf-capability to rbcs … that’s a little more involved, I am afraid.

OK, I'll look at exf and how it works with obcs.  I didn't know exf and obcs could play together.

Thanks,   Jody

> Martin
> 
> On Mar 5, 2014, at 4:48 AM, Jody Klymak <jklymak at uvic.ca> wrote:
> 
>> Hi again,
>> 
>> OK, to answer my own question, no, there is no equivalent of rbcsForcingOffset for obcs.  
>> 
>> Can I suggest in obcs_fields_load.F there be an offset variable, and that this be added to OBCS_PARAMS.h and obcs_readparms.F?  
>> 
>>     CALL GET_PERIODIC_INTERVAL(
>>    O                  iRecP, iRec0, iRec1, bWght, aWght,
>>    I                  externForcingCycle, externForcingPeriod,
>>    I                  deltaTclock, myTime-obcsForcingOffset, myThid )
>> 
>> If you want to use a snapshot, then, it is just a matter of setting 
>> 
>> obcsForcingOffset=-externForcingPeriod/2
>> 
>> Or, one could imagine this being called "externForcingOffset" and being used in "external_fields_load.F" as well....
>> 
>> Cheers,   Jody
>> 
>> 
>> 
>> On Mar 4, 2014, at  17:52 PM, Jody Klymak <jklymak at uvic.ca> wrote:
>> 
>>> One other quick question that I know has come up before but I can't seem to find...
>>> 
>>> I have snapshots at t=0,3600,7200, and I want to use them for my obcs.  However, I seem to remember that externalForcingPeriod=3600 assumes the first file entry is at t=1800, the second at 5400s etc.  Is there an offset to change this so that the first is t=0, etc, or do I have to hack get_periodic_interval.F?  
>>> 
>>> Thanks,   Jody
>>> 
>>> 
>>> On Mar 4, 2014, at  9:18 AM, Jody Klymak <jklymak at uvic.ca> wrote:
>>> 
>>>> Hi Dimitris,
>>>> 
>>>> I wrote you offline as well, but...
>>>> 
>>>> On Mar 4, 2014, at  9:06 AM, Dimitris Menemenlis <dmenemenlis at gmail.com> wrote:
>>>> 
>>>>> Hi Jody, I think yes.  What grid spacing are you running your regional domain?
>>>>> And what resolution (space and time) is your open boundary conditions?
>>>> 
>>>> Right now I have hourly (well 3720 s to be an even divisor of an M2 tidal cycle) output on a 5- to 1-km grid, with 1 km near my region of interest, and 5 km at the boundaries.  The forcing of this model was an analytical source function (a couple of internal tide line-sources outside the domain), and the model was spun up for 12 tidal cycles allowing reflections from a rough continental slope to set up. Probably accurate out to mode 10 or so.  
>>>> 
>>>> I am trying to telescope this down to about 100x200 m in a 50x100 km region to look at the non-linear internal waves on a continental slope.  I nest this inside a 400x300 km region using telescoping resolution, and will initialize and force with the larger 1-km grid. 
>>>> 
>>>> All the telescoping is to help make the BCs not too important, but it would be nice to make them as close as possible to reduce reflections. 
>>>> 
>>>> Thanks!   Jody
>>>> 
>>>>> 
>>>>> Michael Schodlok has a lot of experience going from
>>>>> CS510 (18 km) to 1 km domains, then adding tides,
>>>>> and possibly obcs sponge to get rid of some edge artifacts.
>>>>> 
>>>>> But with sufficiently high-resolution (in time) obcs,
>>>>> I am guessing that you will not need sponging at all.
>>>>> At least that's the theory.
>>>>> Chris Hill was experimenting with "replaying" bits
>>>>> of the 1/48th using obcs.  I don't know how far he got.
>>>>> 
>>>>> Let me know if you want to access the 1/48th output
>>>>> and we can figure out best way to proceed.
>>>>> 
>>>>> Cheers, Dimitris
>>>>> 
>>>>> On Mar 4, 2014, at 8:50 AM, Jody Klymak <jklymak at uvic.ca> wrote:
>>>>> 
>>>>>> Do you think its true that I could just force w/ obcs and do away w/ the sponge? I've used sponges for so long I hadn't thought of giving this a try, but maybe it is all I need to do, and I can ditch rbcs.  Certainly my forcing should have all the modes that could conceivably reach the boundaries by the time my integration is done.  
>>>>>> 
>>>>>> Second, 1/48th degree global with tides!  That would be very interesting!
>>>>> 
>>>>> 
>>>>> _______________________________________________
>>>>> MITgcm-support mailing list
>>>>> MITgcm-support at mitgcm.org
>>>>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>>>> 
>>>> --
>>>> Jody Klymak    
>>>> http://web.uvic.ca/~jklymak/
>>> 
>>> 
>>> _______________________________________________
>>> MITgcm-support mailing list
>>> MITgcm-support at mitgcm.org
>>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>> 
>> 
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-support
> 
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support