Erroneous subdomain output CCLM6 – in #9: CCLM

in #9: CCLM

<p> Hi all, </p> <p> I am currently downscaling ERA5 to EUR11 and the plan is to subsequently downscale this to a higher resolution domain over Belgium. </p> <p> To limit memory usage of the EUR11 run, I decided on using a subdomain, described in out01 with these setting: </p> <p> &amp;GRIBOUT <br/> hcomb         = 0,30000,1, <br/> ytunit        = 'd', <br/> lwrite_const  = .TRUE., <br/> lcheck        = .FALSE., <br/> luvmasspoint  = .FALSE., <br/> ydir          = 'EUR11/cclm/out/out01/', ! nesting variables <br/> yform_write   = 'ncdf', <br/> yvarml        = 'FRESHSNW','HHL','P','PP','QC','QG','QI','QR','QS','QV','QV_S','RHO_SNOW','T','T_S','T_SO','T_SNOW','U','V','W','W_I','W_SNOW','W_SO', <br/> ireset_sums   = 2, <br/> ydomain       = 's', <br/> slon          =  -15.615, <br/> slat          =  -4.455, <br/> elon          =  -1.865, <br/> elat          =  6.105, <br/> /END </p> <p> However, as you can see from the figure below, this does not give the expected output. The output from the run where we do not use a subdomain is fine. </p> <p> Any idea on what the issue/reason could be? </p> <p> Kind regards </p> <p> Fien </p> <p> </p> <p> </p> <figure class="image"> <img src="/uploads/media/material/f8eb3555-f35d-4a95-be65-f85434ae9bf4/T2M_1982010100.png"/> </figure> <p> </p>

  @fienserras in #951a59c

<p> Hi all, </p> <p> I am currently downscaling ERA5 to EUR11 and the plan is to subsequently downscale this to a higher resolution domain over Belgium. </p> <p> To limit memory usage of the EUR11 run, I decided on using a subdomain, described in out01 with these setting: </p> <p> &amp;GRIBOUT <br/> hcomb         = 0,30000,1, <br/> ytunit        = 'd', <br/> lwrite_const  = .TRUE., <br/> lcheck        = .FALSE., <br/> luvmasspoint  = .FALSE., <br/> ydir          = 'EUR11/cclm/out/out01/', ! nesting variables <br/> yform_write   = 'ncdf', <br/> yvarml        = 'FRESHSNW','HHL','P','PP','QC','QG','QI','QR','QS','QV','QV_S','RHO_SNOW','T','T_S','T_SO','T_SNOW','U','V','W','W_I','W_SNOW','W_SO', <br/> ireset_sums   = 2, <br/> ydomain       = 's', <br/> slon          =  -15.615, <br/> slat          =  -4.455, <br/> elon          =  -1.865, <br/> elat          =  6.105, <br/> /END </p> <p> However, as you can see from the figure below, this does not give the expected output. The output from the run where we do not use a subdomain is fine. </p> <p> Any idea on what the issue/reason could be? </p> <p> Kind regards </p> <p> Fien </p> <p> </p> <p> </p> <figure class="image"> <img src="/uploads/media/material/f8eb3555-f35d-4a95-be65-f85434ae9bf4/T2M_1982010100.png"/> </figure> <p> </p>

Erroneous subdomain output CCLM6

FS wrote

Hi all,

I am currently downscaling ERA5 to EUR11 and the plan is to subsequently downscale this to a higher resolution domain over Belgium.

To limit memory usage of the EUR11 run, I decided on using a subdomain, described in out01 with these setting:

&GRIBOUT
hcomb         = 0,30000,1,
ytunit        = 'd',
lwrite_const  = .TRUE.,
lcheck        = .FALSE.,
luvmasspoint  = .FALSE.,
ydir          = 'EUR11/cclm/out/out01/', ! nesting variables
yform_write   = 'ncdf',
yvarml        = 'FRESHSNW','HHL','P','PP','QC','QG','QI','QR','QS','QV','QV_S','RHO_SNOW','T','T_S','T_SO','T_SNOW','U','V','W','W_I','W_SNOW','W_SO',
ireset_sums   = 2,
ydomain       = 's',
slon          =  -15.615,
slat          =  -4.455,
elon          =  -1.865,
elat          =  6.105,
/END

However, as you can see from the figure below, this does not give the expected output. The output from the run where we do not use a subdomain is fine.

Any idea on what the issue/reason could be?

Kind regards

Fien

View in channel
<p> Hello Fien, </p> <p> there really is a bug in src_output.f90 in the new version. Erroneously not the field containing the subdomain has been passed to the routine write_netcdf, but the full field. I added the src_output.f90 (based on the latest COSMO NWP Version) with the bug fix. You can find it in line 5686ff. </p> <p> I hope I uploaded src_output.f90 correctly. It should be in the “All material” </p> <p> Regards </p> <p> Uli </p>

  @ulrichschättler in #6daac74

<p> Hello Fien, </p> <p> there really is a bug in src_output.f90 in the new version. Erroneously not the field containing the subdomain has been passed to the routine write_netcdf, but the full field. I added the src_output.f90 (based on the latest COSMO NWP Version) with the bug fix. You can find it in line 5686ff. </p> <p> I hope I uploaded src_output.f90 correctly. It should be in the “All material” </p> <p> Regards </p> <p> Uli </p>
US replied

Hello Fien,

there really is a bug in src_output.f90 in the new version. Erroneously not the field containing the subdomain has been passed to the routine write_netcdf, but the full field. I added the src_output.f90 (based on the latest COSMO NWP Version) with the bug fix. You can find it in line 5686ff.

I hope I uploaded src_output.f90 correctly. It should be in the “All material”

Regards

Uli

View in channel