mesh subroutine: is_topology_non_periodic bug

### Version

main

### Are there any linked Issues or Pull Requests?

MetOffice/lfric_core#366

### What happened?

The **is_topology_non_periodic** subroutine of the mesh object has a bug, which is fixed in the linked Pull Request MetOffice/lfric_core#383. Therefore, all code calling this subroutine will change results. At the moment, only the **lfric2lfric** app uses it, so all affected KGOs will need to be updated. We will use the opportunity to fix some warnings that appear in the global_variables_checker.

### Relevant log output

```shell

```