Add IFS/HTESSEL grib codes for vegetation related variables

[treerink]

The request from Annalisa, see ECE portal issue 1312-106:

The output coming from the grib codes for the following IFS/HTESSEL variables is requested: 27(cvl), 28(cvh), 29(tvl), 30(tvh), 66(lai_lv), 67(lai_hv) at 6-hrs frequency is fine (as we will need them monthly).

Instead of adding a special list of grib codes with some adhoc special infrastructure for that, it is easier for the genecec system just to add these variables (although there is no need from the side of the requesting people), with that it is also possible to cmorise (including remapping) these variables allowing a quick and easy glance on them.

grib code	longname	CMOR table - variable	based on
27.128	Low vegetation cover (cvl)	HTESSSELmon cvl	Lmon c3PftFrac
28.128	High vegetation cover (cvh)	HTESSSELmon cvh	Lmon c3PftFrac
29.128	Type of low vegetation (tvl)	HTESSSELmon tvl	Lmon c3PftFrac
30.128	Type of high vegetation (tvh)	HTESSSELmon tvh	Lmon c3PftFrac
66.128	Leaf area index, low vegetation (lai_lv)	HTESSSELmon lai_lv	Lmon lai
67.128	Leaf area index, high vegetation (lai_hv)	HTESSSELmon lai_hv	Lmon lai

The variables have not been added to the ACCUMFLD list in grib_codes.json, as I thought they shouldn't.

[treerink]

The updated output-control-files for this request have been commited in the optimesm branch in the ECE repository, see ECE portal post 1312-111 for the comment on the 3 hourly output.

[treerink]

However, there is one small issue to solve I realize now: the conflict of the three add scripts:

add-htessel-vegetation-variables.sh
add-lpjg-cc-diagnostics.sh
add-variables-for-co2box.sh

Currently the latter (the add-variables-for-co2box.sh) does not cause a conflict. But the maintenance friendly & robust approach will cause a conflict currently with the modifications applied to CMIP6_CV.json when the two other scripts are run at the same time at cmorisation stage (this did not affect running genecec).

[treerink]

I post-pone a solution for this for after the back merge of the co2box branch into the master (#792 #785).

[treerink]

Last potential issue solved, one could run now:

./add-lpjg-cc-diagnostics.sh
 ./add-variables-for-co2box.sh
 ./add-htessel-vegetation-variables.sh no-clean-before

[treerink]

Added request and comment from the ECE portal issue 1312-120 post:

The cvl/cvh/lai_lv/lai_hv variables should ideally be saved daily, but monthly is enough for general evaluation.

Addressed in 0082f65.

Cmorisation works fine for lai and cv, but the results for vegetation type (tv) are just rubbish.
(A possible solution could be to replace setgridtype,regular by remapnn for the remapping with cdo)

actually what is needed is setgridtype,regularnn for the tvl/tvh variables

Addressed in 176a74d.

[treerink]

84ea812 fixes a small but essential BUG at cmorisation time.