I gave some thought about future "shapes" of those graphs few years ago, crandles. My guess is, summer (minimum) extent has downward acceleration and will keep having it all the way to (near) zero minimum extent, the reason for this - is consideration of main forces which cause the process.
I mean, let us see, what exactly forces the process of decreasing annual minimum sea ice extents? AFAIK, it is those forces:
- greenhouse gases increase
- warmer water and air entering Arctic from the south
- decreasing (as a result of earlier and larger melts) albedo
- massive amount of sunlight during summer in Arctic (polar day - sunlight 24/7)
Now, as far as i know, all except the 4th - are increasing in power as years go by. CO2 levels are rising with acceleration, methane release in Arctic itself is also accelerating, SSTs in "south" are rising with acceleration, loss of albedo in Arctic and subpolar regions - is also accelerating. The last force - sunlight, - remains nearly constant.
So, in total, forces which cause decreasing annual minimum sea ice cintent in Arctic - are getting stronger not linearly, but with acceleration. And since we know the end result _can_ be zero summer ice in Arctic - i don't see any large reason why the downward acceleration on the graph for minimum annual ice extent could stop. It should remain.
Accordingly, since there can't be much volume if the extent is zero, - the same (downward acceleration) dynamic will remain to be present on minimum _volume_ graphs for Arctic sea ice.
Which means we'll definitely see ice-free summer (late summer, first) in Arctic in a few years. May be 2015 or 2016.
However, maximum extent is not the same. I believe that present "downward acceleration" on maximum winter extent graph - will remain for some time, but the acceleration will at some point start to decrease, and reach 0. The reason for it - is polar night. See, from purely practical, every-day experience we all know that in subpolar regions (places not so far from 60 degrees north latitude), - lakes during summer can get quite warm, but during dark (very little sunlight) winters, - ice forms despite high summer water temperatures.
Good example is lake Baikal. It's a huge mass of water. During summer, with lots of sunlight, surface water temperatures in Baikal are: in the middle of the lake it's 14...15°C (august), in shallow bays - some 18...22°C. Yet, whole lake freezes every year by January, and remain mostly frozen (surface) till May.
I expect the same to happen in much of the Arctic, threfore, maximum (winter) sea ice extent will stabilize at some significant figure. The "downward acceleration" which we can see for maximum ice extent graphs today - will disappear, and then, deceleration trend will start. Further very small long-term reduction in maximum extent will probably be (due to on-going global warming), of course, but i think maximum extent will remain nearly stable.
Maximum volume, though, will keep its downward acceleration for much longer, i guess. It takes very thin ice to count an area as "a part of ice extent", and with ice being lighter than water, thin "skin" of ice can stay on top of much warmer (than -1.9°C, which is frezing point for sea water) water. I think it all starts with snow: without sunlight, athmosphere gets into negative celsius temperatures, snow forms, snow falls to the Arctic ocean, and even if the water is very warm, - melt snowflakes form a very thin layer of low-salinity water, which freezes easier than sea water. This, very thin, surface layer of low-salinity water emits IR radiation and also loses heat by convection (since the athmosphere is colder), which inevitably leads to ice formation. Even if water below it is much warmer. The warmer the water is, the slower is ice growth. Surfaces waves is a big factor when present, too. But, even with waves, it just needs "enough snow" to form thick enough layer of less-salinity (and thus, less-density) water for freezing to eventually overcome the heat of deeper (and possibly much warmer, but also high-salinity and thus more dense) layers, and thus form some thin surface ice. And then it's a matter of equilibrium between how much heat deeper layers are still (slowly) transporting to the surface - which is a force preventing further freezing or even melting some ice from-below, if intense enough, - and how much energy is lost from the ice surface via convection (to a very cold athmosphere) and IR radiation.
So, winter freezing in terms of _volume_ will continue to be slowed and decreased by those 3 accelerating forces mentioned above, and maximum volume will continue to be decreased in more or less linear manner for at least several decades, i think. GHGs are important even during polar night, since they trap heat which Earth emits; albedo loss is still important even without sunlight due to its "delayed" effects - higher heat content of water columns, (shallow) sea floor and shores; air and water currents keep going during winter, some from the south, too.
