[MITgcm-support] Error with Arctic configuration and runoff file

gmail mozzatoale at gmail.com
Thu Feb 18 10:48:13 EST 2016


Hi Jean-Michel,
thank you very much for your email. It was very explicative and I finally managed to solve the problem.
I wanted to interpolate the runoff file as the grids I am using are 210x192, 420x384 and 840x768 whereas the runout files uses a 360x180 grid.
It turns out that the 36km configurations was in fact working with the runoff_interpMethod = 0, setup. Therefore it was working for some reason but it was not actually doing what it was supposed to do!
I have now used the correct runoff interpolation set up for the 18km configuration and it is working fine!
All the best,
Alessandro

> On 18 Feb 2016, at 14:49, Jean-Michel Campin <jmc at ocean.mit.edu> wrote:
> 
> iHi,
> I gorgot one thing:
>>  You can check the size of the file,
> the size of the file I was refering to is just what "ls -l" gives.
> Jean-Michel
> 
> On Thu, Feb 18, 2016 at 09:35:47AM -0500, Jean-Michel Campin wrote:
>> Hi Alessandro,
>> 
>> Sorry, it's not very clear to me: 
>> - do you want to use the model (pkg/exf) to also interpolate the run-off ?
>>  In this case, runoff_interpMethod should be set to 1 (the default) and
>>  you should specify runoff_nlat-1 values for  runoff_lat_inc , like:
>>>  runoff_lat_inc    = 179*1.,
>>>  runoff_nlat       = 180,
>>  Note that this is different from: runoff_lat_inc=179., (only 1 value instead
>>   of 179 values all equal to 1)
>> Or, 
>> - if you don't want the model to interpolate the run-off (I think this is 
>>  what was used in the original set-up you pointed to), you should
>>  explicitly set runoff_interpMethod = 0 and provide a run-off file that
>>  match the model grid (and will change from one resolution to the other).
>>  You can check the size of the file, which I assume should be, with runoffperiod = -12,
>>    Nx*Ny*12*4  with exf_iprec=32
>>    Nx*Ny*12*8  with exf_iprec=64
>>  where Nx,Ny are model grid size from SIZE.h
>> 
>> Cheers,
>> Jean-Michel
>> 
>> On Thu, Feb 18, 2016 at 10:00:41AM +0000, gmail wrote:
>>> Hello everybody,
>>> thank you very much for the suggestions.
>>> 
>>> I have changed runoff_interpMethod  to 1, but this keeps giving me the same "attempt to access non-existent record,??? error
>>> 
>>> However, I have noticed that when I use: 
>>>>>>> runoff_lon0       = 0.50D0,
>>>>>>> runoff_lon_inc    = 1.0D0,
>>>>>>> runoff_lat0       = -89.5D0,
>>>>>>> runoff_lat_inc    = 1.,
>>>>>>> runoff_nlon       = 360,
>>>>>>> runoff_nlat       = 180,
>>> 
>>> both with runoff_lat_inc    = 1., and runoff_lat_inc    = 179., as someone suggested, I get a different error:
>>> 
>>> forrtl: severe (19): invalid reference to variable in NAMELIST input, unit 11, file /tmp/fortxB52Va, line 53, position 12
>>> 
>>> I suppose this is related to the lines I add because without those lines and the runofffile ones the model runs fine.
>>> 
>>> Alessandro
>>> 
>>>> On 17 Feb 2016, at 18:27, Dimitris Menemenlis <dmenemenlis at gmail.com> wrote:
>>>> 
>>>> if you use a 360x180x12 forcing file, then you need:
>>>> 
>>>> &EXF_NML_04
>>>> runoff_interpMethod = 1,
>>>> 
>>>> (which is the default value)
>>>> 
>>>>> On Feb 17, 2016, at 10:19 AM, gmail <mozzatoale at gmail.com <mailto:mozzatoale at gmail.com>> wrote:
>>>>> 
>>>>> Hi matt, 
>>>>> Thanks for the suggestion,  I forgot to mention my data.exf is configured for:
>>>>> runoff_interpMethod = 0,
>>>>> The runoff is indeed a 12 month climatology.
>>>>> In my tests I have tried the configuration you suggest and the same error is thrown.
>>>>> All the best,
>>>>> 
>>>>> Alessandro
>>>>> 
>>>>>> On 17 Feb 2016, at 18:09, Matthew Mazloff <mmazloff at ucsd.edu <mailto:mmazloff at ucsd.edu>> wrote:
>>>>>> 
>>>>>> Hello
>>>>>> 
>>>>>> Is your runoff input a 12 month climatology without need for interpolation. If so you need only:
>>>>>> 
>>>>>> In   &EXF_NML_02
>>>>>> runoffperiod      = -12,
>>>>>> and obviosly set runoffFile
>>>>>> 
>>>>>> and in  &EXF_NML_04
>>>>>> runoff_interpMethod = 0,
>>>>>> 
>>>>>> That works for me...
>>>>>> 
>>>>>> Matt
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>> On Feb 17, 2016, at 9:45 AM, gmail <mozzatoale at gmail.com <mailto:mozzatoale at gmail.com>> wrote:
>>>>>>> 
>>>>>>> Dear MITgcm users,
>>>>>>> 
>>>>>>> I am currently running an Arctic configuration of the MITgcm in 3 different resolutions: 36km, 18km and 9km.
>>>>>>> I started using the ERA  dataset for forcing and all three worked fine. 
>>>>>>> 
>>>>>>> I have now switched to the CORE2 dataset forcing and setup my data.exf following the one I found here:
>>>>>>> http://mitgcm.org/viewvc/MITgcm/MITgcm_contrib/gael/verification/global_oce_llc90/input.core2/ <http://mitgcm.org/viewvc/MITgcm/MITgcm_contrib/gael/verification/global_oce_llc90/input.core2/>
>>>>>>> 
>>>>>>> This new setup works fine on the 36km configuration producing satisfying results.
>>>>>>> The problems arise with the 9km and 18km configuration. More specifically the problem seems to be related to the runoff.
>>>>>>> Both configurations throw the following error:
>>>>>>> 
>>>>>>> forrtl: severe (36): attempt to access non-existent record, unit 9, file /hpcdata/scratch/am8e13/CORE_data/runoff-corev2_360x180x12.bin
>>>>>>> 
>>>>>>> Here is a copy of my data.exf:
>>>>>>> https://github.com/AlessandroMozzato/notebooks/blob/master/data.exf <https://github.com/AlessandroMozzato/notebooks/blob/master/data.exf> 
>>>>>>> 
>>>>>>> I have checked the file in MATLAB and it seems fine and with the right dimensions.
>>>>>>> I have also tried to substitute the runoff from the CORE dataset with the runoff from the ERA dataset and the same error is produced. 
>>>>>>> I have tried both of these setups for the data.exf:
>>>>>>> 
>>>>>>> runoffstartdate1  = 19470115,
>>>>>>> runoffstartdate2  = 120000,
>>>>>>> runoffperiod      = 2628000.0,
>>>>>>> runoffperiod      = 2592000.0,
>>>>>>> 
>>>>>>> and:
>>>>>>> 
>>>>>>> runoffperiod      = -12,
>>>>>>> 
>>>>>>> Finally I have added the following lines to the data.exf
>>>>>>> 
>>>>>>> runoff_lon0       = 0.50D0,
>>>>>>> runoff_lon_inc    = 1.0D0,
>>>>>>> runoff_lat0       = -89.5D0,
>>>>>>> runoff_lat_inc    = 1.,
>>>>>>> runoff_nlon       = 360,
>>>>>>> runoff_nlat       = 180,
>>>>>>> 
>>>>>>> It is also worth noticing that removing the runoff file from data.exf solves the problem and the model runs fine. 
>>>>>>> I suspect this could be an easy fix but I couldn???t figure it how.
>>>>>>> 
>>>>>>> Every suggestion would be greatly appreciated!
>>>>>>> 
>>>>>>> All the best,
>>>>>>> Alessandro
>>>> 
>>>> 
>>>> _______________________________________________
>>>> 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
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support




More information about the MITgcm-support mailing list