w3iors NSEAL vs NSEA in non-MPI portions of read/write

[JessicaMeixner-NOAA]

Describe the bug
Not 100% sure this is a bug, but this should be investigated. While working on #829

I found that in the non-MPI sections of the read/write of the restart we are using:

DO JSEA=1, NSEAL

In the following lines of code:
https://github.com/NOAA-EMC/WW3/blob/develop/model/src/w3iorsmd.F90#L631
and
https://github.com/NOAA-EMC/WW3/blob/develop/model/src/w3iorsmd.F90#L788

In standalone mode, perhaps we've already defined NSEAL=NSEA and this is fine, but I found I had to set NSEAL=NSEA when trying to read/write this from ww3_gint.

To test if this is a bug or not, my plan is to:

• run the regression tests with these NSEAL->NSEA and see if they work
• See if there are differences, and make sure to test a non-mpi case for a restart run and make sure everything is working as expected.

[MatthewMasarik-NOAA]

This does sound like a good thing to investigate. Good find @JessicaMeixner-NOAA

[JessicaMeixner-NOAA]

All regression tests except the usual passed. Next step is running a restart case for non-mpi with the develop and this branch and see how that goes.

[ukmo-ccbunney]

So - my opinion on this is that it aught to be changed to NSEA.

Using NSEAL makes the assumption that setting NSEAL = NSEA is always handled by whatever code is calling this.
In the wave model, this is always the case (handled via w3parall.f90:SET_UP_NSEAL_NSEALM), but as in this case it is not always the wave model calling this module so we should not make assumptions. NSEA will always be defined, but AFAIK NSEAL is only ever valid in an MPI case so we should no expect it to be defined by the calling routine in SHRD mode.

Assuming that changing it does not impact the results (I can't see why), then I advocate defensive programming and changing it to NSEA.

There might be other places in the code where this assumption on NSEAL being defined is also made...

[JessicaMeixner-NOAA]

@ukmo-ccbunney okay -- I will proceed with this code-fix. Should have a PR up soon... that's a good question about other places in the code about the assumption on NSEAL. Should we make a new issue for that?

[MatthewMasarik-NOAA]

I find @ukmo-ccbunney suggestion compelling.

[ukmo-ccbunney]

@ukmo-ccbunney okay -- I will proceed with this code-fix. Should have a PR up soon... that's a good question about other places in the code about the assumption on NSEAL. Should we make a new issue for that?

Perhaps not a bad idea so it is logged and someone can then put some time aside to look through the code.

[ukmo-ccbunney]

I'm running regression tests now.

[JessicaMeixner-NOAA]

New issue: #956 for general case