BLOM density calculation CGS versus MKS at initialization

[jmaerz]

Dear @matsbn and @milicak , when running the single column setup with -D mks=true -D openmp=disabled there is written out init.profile at initialization in the log file which hold very different density values, 91.97 versus 31.04 with cgs, while temperature and salinity look the same. Could you check? - I am not very familiar with the BLOM internals yet.