add BLOM regression test functionality

[mvertens]

This PR has changes to buildnml and to the testdefs directory to enable BLOM to have its first set of regression tests.
This PR resolves issue #258.

To see how the regression test is working - do the following:
(Note: the BLOM external points to my repository until this PR is merged)

> git clone https://github.com/mvertens/NorESM.git
> cd NorESM
> git checkout noresm_develop_v7_proto
> cd cime/scripts
> ./query_testlists --xml-category aux_blom_noresm

will give the following output

aux_blom_noresm: SMS_D_Ld1.T62_tn14.NOINYOC.betzy_intel
aux_blom_noresm: ERS_Ld5.T62_tn14.NOINYOC.betzy_intel
aux_blom_noresm: ERS_Ld5.T62_tn14.NOINY.betzy_intel
aux_blom_noresm: ERS_Ld5.T62_tn14_wtn14nw.NOINY_WW3.betzy_intel
aux_blom_noresm: ERS_Ld5.T62_tn14_wtn14nw.NOINY_WW3.betzy_intel.blom-wavice

To run a test suite and also generate baselines:

./create_test 
    --xml-category aux_blom_noresm  
    --xml-machine betzy --xml-compiler intel 
    --generate <your_baseline_directory_name>
    --baseline-root /cluster/shared/noresm/noresm_baselines 
    --project <your_project>

Baselines for the above aux_blom_noresm tests will be created in
/cluster/shared/noresm/noresm_baselines/<your_baseline_directory_name>

More documentation on regression testing in NorESM can be found at:
https://noresm-docs.readthedocs.io/en/noresm2/testing/systemtests.html

[TomasTorsvik]

I should look more into the regression test functionality, but I don't see any reason to hold back on merging this PR.

[mvertens]

@TomasTorsvik - thanks for the review. Do we need to wait for the review for Mats as well. I'd like to get this into a updated NORESM development snapshot that is needed for the Making Waves project. This is the last branch that is still pointing to my own forked code.

[TomasTorsvik]

@TomasTorsvik - thanks for the review. Do we need to wait for the review for Mats as well. I'd like to get this into a updated NORESM development snapshot that is needed for the Making Waves project. This is the last branch that is still pointing to my own forked code.

@mvertens - I usually wait for review from Mats when a PR touches physics or model coupling (i.e. not purely BGC). Since this PR is mainly adding test functionality it is different from the PRs that we usually deal with. If you need this functionality in your current work, I suggest we merge it so that it doesn't stop you in your progress. Otherwise I would prefer to wait for input from Mats.