At this point, I'd like to point out a remarkable consistency between IMB (Ice Mass Balance) buoy 2017A and co-located WARM buoy 6 when it comes to ice bottom melt.
IMB buoy 2017A shows this melting profile :
Note that over July, that amounts to about 2 cm/day of bottom melt.
WARM buoy 6 records this profile of in-ice and under-ice PAR irradiance :
which suggests PAR irradiance (light between 400 nm and 700 nm) was some 30 W/m^2 over July.
We know that sunlight holds some 43% of its energy between 400 and 700 nm, so we can reasonably expect that some 30 * 1/0.43=70 W/m^2 of sunlight made it through the ice in July.
70 W/m^2 is enough heat to melt 2 cm/day of ice, very consistent with the 2 cm/day we see recorded by 2017A. Other WARM buoys (such a WARM 3) suggest that this level of irradiance is common, and lasts July through Aug.
This is telling although not quite unexpected. It suggests that light that makes it through the ice is THE main source of bottom melt in the Arctic sea ice pack.
It also tells a lesson, since this should be happening all through the Arctic : Even without external influence, simply the Arctic light that shines through the ice can account for 2 cm/day bottom melt through July and August. That's 120 cm right there. Add 30 cm of top melt and we can conclude that any ice that is 1.5 meter of ice thickness or less at the start of the melting season is at risk of melting out in-situ.