Welcome, Guest
Username: Password: Remember me
Forum header

TOPIC: fldext.f question

fldext.f question 9 years 6 months ago #42

  • Ulf Andrae
  • Ulf Andrae's Avatar
  • OFFLINE
  • Administrator
  • Posts: 288
  • Thank you received: 32
I'm assessing the impact of ASCAT observations assimilation on Hirlam. I intended to use the vfld files which are created by
/home/ms/dk/nhz/harmonie_release/trunk/util/gl/prg/fldextr.F
on ECMWF.

As the largest impact of ASCAT assimilation is expected over sea I planned to use the Vfld output for moored buoys. Unfortunately I fail to ensure that the moored bouys information is appearing in the vfld output files created by fldextr.F.

I incorporated the moored buoys into the allsynops.list which is copied by the VER_create_fld script to synop.list. This did not lead to the desired result : vfld files containing moored buoys.

I assume that it has something to do with the land_limit=0.01 setting in the VER_create_fld script( but I'm not sure about this).
I couldn't find information on this on the Hirlam Twiki page. Only that some modifications were done in 2007.
How can I get sea surface data into the vfld files, or it this not possible?

Re:fldext.f question 9 years 6 months ago #43

  • Ulf Andrae
  • Ulf Andrae's Avatar
  • OFFLINE
  • Administrator
  • Posts: 288
  • Thank you received: 32
The way to turn off the land sea mask check in fldextr is as indicated to set land_limit < 0 in VER_create_fld. If you check the reading routine for e.g. hirlam data, getgrb_any.f90, you will find that the mask for which gridpoints to use will be set identical TRUE if land_limit < 0.

Please report back if it doesn't work.

Regards,

Ulf

Re:fldext.f question 9 years 6 months ago #44

  • Paul de Valk
  • Paul de Valk's Avatar
Just to be absolutely sure:
Do I interpret this correctly that any value of land_limit below zero is sufficient? -0.01 and -1.0 will have the same effect?

Regards
Paul

Re:fldext.f question 9 years 6 months ago #45

  • Ulf Andrae
  • Ulf Andrae's Avatar
  • OFFLINE
  • Administrator
  • Posts: 288
  • Thank you received: 32
Strictly speaking the comparison is done on the land sea mask like:
mask = ( lsm >= land_limit )

where mask is a logical field with the same dimension as lsm. So you have to make sure that land_limit is less than your smallest possible land sea mask value ( which should be zero).

Ulf

Re:fldext.f question 9 years 5 months ago #46

  • Paul de Valk
  • Paul de Valk's Avatar
Dear Ulf,

Thanks for your suggestion. It works! The required moored buoys appear in the vfld files.
I'm nearly there where I want to be. When I also can get this file list into the vobs files i can start the analyses.

Best regards Paul
Time to create page: 0.085 seconds