The ice is warmer than the melting point of seawater - when the ice heats up in the summer, the heat conducts down to the base where it melts - the fresher water in the ice mixing with the saltier water below and melting. The simplified thermodynamics are described below. In the real world, the dynamics of melting at the base of the ice are really complicated, particularly as the fresher warmer melt insulates the base of the ice from the higher salinity water below
Consider the surface of the ice is at 0 °C during the summer, and the base of the ice at -1.8°C. The ice has a temperature gradient; the middle of the ice is at -0.9°C. The heat conducts down from the surface. The rate of bottom melt depends on the conductivity of the ice, the thickness, and the transfer of heat to the seawater (assuming it's held at a constant salinity and temp). Thin ice has more bottom melt as it has to conduct heat less far. If the surface of the ice has no surface melting and is held at 0°C, all the ice would still melt out from the bottom melt.
When the temperature at the surface falls to, say, -5°C, the heat from the ice middle at -0.9°C transfers upwards and downwards and cools until it reaches -3.4°C.
Ditto thanks for that explanation Rox. Describing the fact that as the ASI (on average) gets thinner, the required temperature to melt ASI (on average) shifts to a lower threshold illuminates a new perspective in my dim mind.
A 0.5C (pick any number you like) lowering of the average melt temperature threshold (e.g. from ice melting at -1.8C instead of having to reach -1.3C to cause melting) would have the same effect as a +0.5C rise in Arctic Ocean water/surface air temperatures would have had if the ASI had remained the same.
With the combination of a warming Arctic
AND thinning sea ice, the two trends combine to cause melt at a rate greater than either effect (thinning, temperature rise) would have alone.
Is this logic correct?
Nothing new in saying that as the ASI thins it is less resistant to melt. But your explanation of the energy transfer change for thinner ice highlights this double whammy concept.
I guess there's nothing new in that either, except to my brain it highlights how the ASI is in peril from essentially getting hit from two forcings at once as both trends continue on their respective trajectories.
This perspective also brings insight into Peter Wadham's "Just goes Poof" description of the end-game scenario. It seems that the effect of thinning on the melt temperature threshold is relatively minor and very slowly incremental as the ice transitions from 2.0 meter to 1.5 to 1.0 to 0.5 M. As as the ice is >0.5 M thickness it is the continuation and gradual increase of elevated Arctic temperatures that drive the change in the ASI statistics (Extent, Area, Thickness, Volume). While extremely rapid in a geo-paleoclimatic frame of reference, from the perspective of an ASIF fan watching from the cheap seats, it progresses relatively slowly from year to year.
But another force emerges to accelerate the pace as thickness gets below about 0.5 M. At that late stage the formerly rather insignificant cumulative effects from the decline in melt temperature threshold, the distance heat has to travel through the ice, and in the amount of energy required to melt the remaining thickness (from a combination of the amount of ice in that thickness and its temperature dynamics per unit) will become increasingly more important in determining the ice melt response within a single melt season.
Because of the acceleration provided by those late-stage additional melt rate drivers, the time to go from 2.0 to 1.5 to 1.0 to 0.5M will be on one trajectory, but that final 0.5M, and especially the last 0.25 M, would be on a much steeper and faster trajectory. As a result, removing the final 0.5M will take a lot less time than it took to go from 2.0 to 1.5M average thickness.
Ice scientists have probably shown this in a chart somewhere, but all I could find in a superficial search was one chart that was really about a broader question and just happened to include a few data points showing increased melt rate as ice thinned down to about 0.5M.
Note that in the 1st chart shown below, which is the data source for the 2nd chart, the negative "Growth Rate" in summer = summer Melt rate. The X-axis on the chart is meters of ASI thickness. --- Yes, Virginia, there is a Santa Claus. And at one time he and Mrs. Claus relaxed during the summer on 15 Meter thick ASI. Now they have to live on a houseboat.
Chew on the fact that not that long ago there was 15M thick summer sea ice in the Arctic. And the chart in the Zhang and Rothrock source article has this annotation: "The ice growth rates for ice thicker than 15 m are not plotted." So they had observations for SUMMER Arctic sea ice THICKER than 15 meters! Now there is precious little ASI thicker than 4M.
The 2nd chart just highlights the melt rate for a few of the thinnest ice data points from the 1st chart. The data points in the 2nd chart are from the upper solid line of the 1st chart, the lower dotted and dashed lines in the 1st chart are irrelevant for this discussion.