Rethinking mode splitting, splitting in general, Boussinesq, non-Boussinesq, seawater EOS, and how it all comes together

A convoluted story about how making code work comes long before understanding of how it works, and understanding why it works leads to re-discovery of old knowledge.  Essentially a unified approach to mathematical splitting of stiff operators, not just for barotropic-baroclinic mode, but overall throughout the oceanic solver. A perturbational analysis helps to derive two- or three-way splits involving several components of the model.  Barotropic mode, compressible EOS, implicit bottom drag can peacefully co-exist without overwriting computational efforts of each other.