Temperature flux due to frazil ice melting in seawater is incorrect

[cbegeman]

When frazil ice melts in MPAS-Ocean it is a freshwater volume flux (not a virtual flux). As such, there should be two components to the temperature tendency due to frazil melt, a latent heat flux and a flux associated with the mixing of water at the freezing point with the ambient seawater. The latter is not included (a bug).

[cbegeman]

The first temperature tendency term that is currently included in master is here:

E3SM/components/mpas-ocean/src/shared/mpas_ocn_frazil_forcing.F

Lines 597 to 600 in 90d3f90

	! ocean fluid temperature is cooled due to melting of frazil
	frazilTemperatureTendency(k,iCell) = - ( meltedFrazilIceThickness * config_frazil_heat_of_fusion &
	* config_frazil_ice_density ) / (config_specific_heat_sea_water * rho_sw) &
	/ dt

The missing term added by #6976 is:

E3SM/components/mpas-ocean/src/shared/mpas_ocn_frazil_forcing.F

Lines 602 to 604 in 90d3f90

	! ocean fluid temperature is modified due to mixing of existing water with meltwater at the freezing point
	frazilTemperatureTendency(k,iCell) = frazilTemperatureTendency(k,iCell) + &
	meltedFrazilIceThickness * oceanFreezingTemperature / dt