FPv10 (2016-11-2) not working with old ECMWF GRIB1 as it was in Summer 2016

[donaldjmorton]

In Summer 2016, I used the latest FPv10 to use two of the old 92-level ECMWF GRIB1 files in test case, and it worked fine:

$ ./FP
 Welcome to FLEXPART Version 10.0beta (2015-05-01)
 FLEXPART is free software released under the GNU General Public License.
 Vertical levels in ECMWF data:      92     92

 Mother domain:
  Longitude range: -179.00000 to  181.00000   Grid distance:    1.00000
  Latitude range :  -90.00000 to   90.00000   Grid distance:    1.00000
.
.
.
 CONGRATULATIONS: YOU HAVE SUCCESSFULLY COMPLETED A FLEXPART MODEL RUN!

Now, in the same test setup, I use the newest FPv10, and I get:

$ ./FPnew
 Welcome to FLEXPART Version 10.1beta (2016-11-02)
 FLEXPART is free software released under the GNU General Public License.
 Vertical levels in ECMWF data:      92     92

 Mother domain:
  Longitude range:  180.00000 to  540.00000   Grid distance:    1.00000
  Latitude range :  -90.00000 to   90.00000   Grid distance:    1.00000

  -179.000000      -90.0000000    
   540.000000       90.0000000       180.000000      -90.0000000       181.000000       90.0000000       1.00000000       1.00000000    
  #### FLEXPART MODEL ERROR! PART OF OUTPUT    ####
  #### GRID IS OUTSIDE MODEL DOMAIN. CHANGE    ####
  #### FILE OUTGRID IN DIRECTORY               ####

I am currently investigating the parallel computing features of this code, and may investigate this problem in more detail.

[donaldjmorton]

OK, this was a quick ticket. It turns out that nxshift (from ecmwf_mod.f90) in my older case had been set to 0, but had been set to 359 in the newer code. I set it back to 0 and all is fine.

Whoever has the authority to close this may close it now.

[pasko]

You can close this ticket yourself by setting the ticket to 'resolve as fixed'