The flip side of this occurs during the melting season when only the ice side of the interface is relevant as the water/brine below is already in liquid state. Here, the melting point is 0.0C at normal atmospheric pressure or 760 torr. Small deviations in pressure as the water gets deeper will result in slightly different melting points.
The brine below the ice controls at what temperatures bottom melt starts.
The brine within the ice controls at what temperatures melting starts within the ice and at the surface of the ice.
As long as bottom melt is active, the temperature is pegged at the melting point of the ice/water interface - which as oren pointed out is at -1.8, since any expelled brine from the freezing process quickly dissipates.
Thermal radiation enters the ice from above and in new and first-year ice, brine pockets within the ice warm up at the same rate as the ice itself. As soon as the ice starts melting at the interfaces of the brine pockets and the surrounding ice, the temperature STOPS RISING. All the extra heat (energy) is used to melt the ice.
Since the brine pockets can quite easily be saltier than the underlying ocean, the resulting starting melting point could quite well be as low as -3 degrees or even lower. As melting progresses, the brine pockets increase in size and their salinity goes down, raising the melting point.
BUT the temperature within the ice never goes above the melting point of the active melting interface. So if the ice contains brine pockets, the temperature of the ice never reaches 0C. Which makes it very obviously wrong to say that the melting point of that ice is 0C.
What about the surface, I hear you ask. We've covered bottom melt and "internal" melt, which is what happens as a result of thermal radiation (i.e. sunlight). Melting caused by conduction from air takes place at the surface, but again, the brine pockets reach all the way to the surface. So the melting ice will always be interacting with brine, and thus melting of the surface will start when air temperatures are at the melting point of the brine-pocket/ice interface.
So all in all, Phoenix, new sea ice has a melting point well below 0, first year ice also has a melting point below 0 but not by much, and multiyear ice has a melting point at 0.
The reason that you will not find any direct information online about the melting point of sea ice is that it is never simple nor obvious (as opposed to the freezing point of sea water).