Check necessity of forcing temperature of element with water to melt temperature
After rev. 183, the flag setting for SubSurfaceMelt/Freeze moved to PhaseChange and is now executed after the computation of snow temperatures.
Previously it was done, before executing the last iteration in the computation of snow temperatures, which may cause differences. Is it now still necessary to force element temperature with water to melt temperature? Should be checked and removed when necessary, as it has the potential of introducing violations of the energy balance.
Note: there is still a remnant of this in compSnowTemperatures (see sn_ElementKtMatrix).