Mechanical tipping points - are we seeing one now?

[Whit]

First time poster here. I decided to jump straight in.

I've been following the blog since last season, while becoming a more and more avid ice-watcher since I started noticing the changes around 10-15 years ago. I'm a complete amateur when it comes to the science of ice, oceans and hydrology.

I hang around in the sub-arctic quite a lot, and the changes in wildlife, vegetation, natural cycles of insects, spring run-off etc has changed tremendously in the areas i frequent. I was born and grew up between 69N and 78N. That background has given me the opportunity to experience a lot of ice-offs, both in fresh and salt water.

Looking at the satellite images and the cracking I'm reminded of what it's usually like when you wait for ice-off on an alpine lake north of the polar circle.

When the ice cracks and takes on a more porous character one early morning, it's usually completely gone the next day. Ice-fishermen have been forced to swim to shore after pitching their tent on what they thought was solid ice.

I believe it has as much to do with mechanics as with how much thermal energy is available to melt ice. What happens IMHO, at least in fresh water, is that as the ice cracks, the surface that water or air can transfer heat to increases exponentially. It's not just the cracks per se. It's also the way the ice turns porous at the onset of melt. This porosity again increases the effective surface of the ice by several orders of magnitude. The porosity also translates to lower albedo.

My gut feeling is that the mechanical and thermodynamic interaction of ice and water, and the way the quality and character of ice changes when at the onset of melt, is underestimated in the models. In particular the extreme increase in effective surface area when the ice first cracks up and then as porosity increases. In addition there is increased surface area for kinetic energy to work, higher drift speed translating to even higher kinetic energy, less resistance towards mechanical forces like wind and waves and so on.

If I'm right 2014 or 2015 at the most seems like a good bet for ice-off considering the current state of the ice, and considering the current trend. Please take into account that ice-off is not the same as ice-free. Ice-off is the last phase before an ice-free state, at least in my vocabulary

This isn't based on much besides personal experience with ice, and I realize the limitations of using freshwater high alpine lakes in the arctic as models of sea-ice. Any input would be much appreciated. I have already learned a lot from the ASIB and ASIG, and now is as good time as any to join the eminent ASIF

[Nightvid Cole]

I think that such lakes certainly have something to tell us, although exactly how similar their mechanical-thermodynamic coupling properties are to sea ice in the Arctic is not obvious.

Perhaps the models of sea ice should be run on lake ice as a sanity check, with, of course, the attendant absence of salinity, lack of a halocline, and lower wind and tidal effects fed in. If nothing else, this might help constrain any parameters not otherwise measurable in the realm of more fundamental physics.

Maybe it would be better to do it on semi-enclosed ocean areas such as Hudson Bay and Baffin Bay, or even a fully enclosed body of saltwater such as the Aral Sea which freezes and thaws every year. For obvious reasons these are better models for the Arctic Ocean than small freshwater lakes. How similar is ice-out on large enclosed or semi-enclosed saltwater bodies (such as those just mentioned) to ice-out on small freshwater lakes?

[LurkyMcLurkerson]

Thanks for the thoughts on it. I am also an amateur when it comes to this science, and though it sounds like you have much more frequent experience with it than I do, I also echo your wonderings. I've seen lakes go. The ice is there, seems fine, and then something mechanical goes on within the structure, and then it is gone. Often, yes, by the next day, depending.

Some of that, I suspect -- from what chemistry background I have, which is decent but by no means huge -- one would need to look at on a much more micro scale than we have many means to do here, or necessarily the backgrounds needed. Any crystal doesn't behave the same when it cracks and reforms as it would have without the crack, the molecular structure of it applies there. Those bonds are extremely persnickety. They aren't seamless from there, and that does effect the energy that it takes to dissociate the parts.

But my knowledge of sea ice in particular is miniscule, so I mostly have wonderings, not any conclusions. As in: what is the state of the molecular structure of the ice, and how does that effect the ways that any energy influx will change its physical form? Much more complex than simple temp, etc, and things for which I have no answer, nor any data whatsoever, so far as I'm aware.

We are seeing things on a macro plane, but the large scale is made up of the molecular interactions for which we have little observational evidence. This is a lot of why I don't trust any of the data enough to trust statistical manipulations on it telling us much. Or, I guess, I trust the data itself to be telling us what it can based on its own bounds, but I don't trust that we're getting the picture we need anymore. Maybe it still is giving us all we need to know, really, but I wouldn't bet anything on it without a clearer picture of the underlying structure beyond "how much ice is there."

Experiences with ice, experiences with changes happening otherwise around those parts of the world -- those are valuable to me, at least, and thank you for sharing them. If anecdote isn't data, it also isn't nothing; it hones our thinking about what the data is, what it isn't, and what questions we should ask, IMO.

This has been a good season for questions. Less so for answers, so far.

[Espen]

Whit,

I can only agree with your observations, and I don't see it is much different with sea water (salty), I have argued for the ice "free" over nite scenario for years now, and I believe in it more now than before.

[Whit]

Thanks for the replies. This weird season just keeps giving and I must admit I was slightly surprised by the slow onset. Nightvid, modelling based on a smaller enclosed body of sea-ice seems like a good idea. Perhaps trying different scenarios with different kinetics.

The Icefjord on Spitsbergen rarely freezes over completely, but I got to see one ice-off there. It was in the late nineties IIRC. First the sun worked the ice for a while, weakening it. And when the winds where right it all disintegrated while hitching a ride towards the open sea. You see the same in lakes, although wind seems to influence ice-off slightly less in smaller bodies of water. That's a generalization. Topography matters a lot it seems.

Espen, yeah. To me it seems completely counterintuitive that ice disappears in a slow linear fashion as predicted by the standard models. I can see a tail of sorts in areas where FYI is compacted, but given the right conditions I feel quite certain that we will get below 1 mill km2 in a stunningly rapid series of events. Most certainly not this year, but if the perennial ice is going (and that's almost a given now) it should go as other forms of floating ice goes - in a flash, when the conditions are right. I believe it could have happened this year. At the same time I haven't written of the possibility for a new minimum record this year. The fragmentation event and the small size of the floes should add to both area and extent. Slush covers more than icebergs ... Whether it makes the ice more resistant to melt is another story.

Barbers study in Beaufort is right on the ball IMO. He states that the new regime of rotten ice should be incorporated in the models. From that I infer that it isn't already. And to me that seems like a plausible explanation for the models complete lack of ability when it comes to predicting the speed of ice-loss we have been seeing.

Waynes work has been most illuminating. I think there might be something to learn from the the ratio between frozen ice and the melt/freeze zone you get in spring. At least for me. You guys might have covered it already

[Peter Ellis]

The process of ice melt within a given year and the year-on-year reduction in the amount of ice remaining at the summer minimum are completely different physical processes!

[Whit]

The process of ice melt within a given year and the year-on-year reduction in the amount of ice remaining at the summer minimum are completely different physical processes!

Certainly. I don't think anyone has said anything contrary to that either. If I have said something that could be interpreted that way, it certainly wasn't my intention.

However, there are similarities. In a small system, like an alpine lake, you have day and night variations with substantial refreeze during the night in the beginning of the melting season. So you get variations around a negative trend, just like we see year-to-year in the Arctic.

Ice that undergoes a melt/freeze cycle is weaker and warmer than ice that just stays frozen. As the ice thins the ratio of the active melt/refreeze zone vs. the frozen zone decreases, thereby decreasing tensile strength of the ice-column by shifting the ratio of weak vs. strong ice.

Exponential loss of ice can be inferred by these and other factors, like volume to surface area ratio and ice-quality, that are quite similar in small and large systems. And as we have seen, the degree of loss has not been modeled very well. Speculating about why is interesting, and other bodies of ice might just provide a clue as to what is happening.

At least that's my 5 cents, and I don't think it's all conjecture either. Our eyes in the sky are amazing tools, but sometimes you actually have to get out there to get the full picture. Extent and area c an only tell you so much. I'll just quote Barber et al 2009; Perennial pack ice in the southern Beaufort Sea was not as it appeared in the summer of 2009:

[1] In September 2009 we observed a much different sea icescape in the Southern Beaufort Sea than anticipated, based on remotely sensed products. Radarsat derived ice charts predicted 7 to 9 tenths multi-year (MY) or thick first-year (FY) sea ice throughout most of the Southern Beaufort Sea in the deep water of the Canada Basin. In situ observations found heavily decayed, very small remnant MY and FY floes interspersed with new ice between floes, in melt ponds, thaw holes and growing over negative freeboard older ice. This icescape contained approximately 25% open water, predominantly distributed in between floes or in thaw holes connected to the ocean below. Although this rotten ice regime was quite different that the expected MY regime in terms of ice volume and strength, their near-surface physical properties were found to be sufficiently alike that their radiometric and scattering characteristics were almost identical.

[Jim Hunt]

A quick note to point out that there is a paywall free version of Barber et. al. 2009 available here.

[Whit]

There it was Than you!