Bugfix in subroutine init_atm_case_mtn_wave regarding parallel execution

[MHBalsmeier]

This is a fix for issue #1256

The problem was that the center coordinate of the domain xc was not synchronized among processes which resulted in wrong orography. This is now fixed and the initial state files created in parallel and serially are identical.

[mgduda]

@MHBalsmeier Thanks very much for finding this bug and providing a fix. I have just three small requests:

1. Rather than opening a PR from your develop branch, could you commit the fix to a new branch based on the develop branch and open the PR from that branch?
2. In the new PR, could you target the hotfix-v8.2.3 branch?
3. In the commit message, could you include an explanation of the bug and how the changes in the commit fix that bug? We use the output of git log quite often, and so more detail in the commit messages can be valuable.

Thanks again, and if you have any questions, please don't hesitate to follow up!

[MHBalsmeier]

Thank you, sure, I'll open a new one.

[MHBalsmeier]

@mgduda "commit the fix to a new branch based on the develop branch and open the PR from that branch": I based my new PR on the current develop branch of this repo, and not on my develop branch ... so now it would merge all the commits in develop into the hotfix branch. It's showing 110 commits. Is that what you intended?
I might have to try a third time, sorry for the chaos.

[mgduda]

@MHBalsmeier I think you'll be able to fix up PR #1312 by force-pushing changes to your existing init_atm_cases_bugfix branch. So, there's probably not a need to make a third PR.

[MHBalsmeier]

@mgduda Thank you, let's see how it looks now.