Flux divergence sign convention (currently inverted)

[btobers]

It seems that the flux divergence has an inverted sign than it should.

I believe that line 1350 of oggm.core.flowline.py should read:

ds['climatic_mb'].data[i, :] - ds['dhdt'].data[i, :]

instead of:

ds['dhdt'].data[i, :] - ds['climatic_mb'].data[i, :]

per Cogley et al. (2010)'s Glossary of Glacier Mass Balance and Related Terms.