Bugfix/st4wcf

[mickaelaccensi]

description :
bugfix for ST4 b4b reprodctibility
linked to issue #249

model :
w3iogrmd : write/read variables AIRCMIN and AIRGB in mod_def.ww3
w3src4md : initialize array
update VERXXX in w3iogrmd

matrix :
done with intel-comp/18, impi/2018.1.163, netCDF-Fortran 4.5.2
ww3_ts1 : results are correct now
ww3_ta1 : results are correct now
ww3_tp2.14 : differences due to metadata history in remap file (not relevant)
ww3_tp2.18 : now differences due to output messages order (dedicated bugfix in progress)
mww3_test_03 : expected not b4b results
mww3_test_04 : different dates in metadata 'history' (it will be skipped in another bugfix branch)

matrix_ST4wcf.zip

[ukmo-ccbunney]

Great work Mickael - well done on finding that bug.
I'll approve once you have the full regression matrix results.

[mickaelaccensi]

I've added the matrix results

[ukmo-ccbunney]

ww3_ta1 and ww3_ts1 are still showing differences in the out_grd.ww3 files (which is expected if you are comparing against the original develop branch, of course).

Just to confirm - are these tests now self consistent with each other?

[mickaelaccensi]

I've compared the figures against the ones from the develop branch in April 21st (before st4tail merge) and I get the similar plots some differences due to the new parameterization. Does this reply to your question ? (it's also now b4b reproducible on my side)

[ukmo-ccbunney]

I've compared the figures against the ones from the develop branch in April 21st (before st4tail merge) and I get the similar plots some differences due to the new parameterization. Does this reply to your question ? (it's also now b4b reproducible on my side)

Yes - thank you Mickael.

[ukmo-ccbunney]

All good to go!

[aliabdolali]

Thanks @mickaelaccensi
I checked the regtests on NOAA HPC with intel, impi and netcdf
matrix_bugfix_ST4wcf.zip