Recovery mechanisms of Arctic summer sea ice Tietsche et al 2011 Discussion

[KiwiGriff]

Warming of the interior Arctic Ocean linked to sea ice losses at the basin margins
Abstract

Arctic Ocean measurements reveal a near doubling of ocean heat content relative to the freezing temperature in the Beaufort Gyre halocline over the past three decades (1987–2017). This warming is linked to anomalous solar heating of surface waters in the northern Chukchi Sea, a main entryway for halocline waters to join the interior Beaufort Gyre. Summer solar heat absorption by the surface waters has increased fivefold over the same time period, chiefly because of reduced sea ice coverage. It is shown that the solar heating, considered together with subduction rates of surface water in this region, is sufficient to account for the observed halocline warming. Heat absorption at the basin margins and its subsequent accumulation in the ocean interior, therefore, have consequences for Beaufort Gyre sea ice beyond the summer season.

https://advances.sciencemag.org/content/4/8/eaat6773.full

[crandles]

Btw hysteresis, really literally means delay! You both refer to a tipping point , a non linear  bifurcation.

Yes, but is that the sense used by IPCC in saying ""there is no indication of hysteresis behavior of Arctic sea ice"?

Clearly increase GHGs and the ocean takes 3 or 4 hundred years to fully come into equilibrium with a higher level of GHGs and the higher ocean temperatures are likely to affect level of sea ice. So clearly there is a long delay before the full effects on sea ice are observed.

So what interpretation should we put on what the IPCC are saying?

I now find the quote saying

A substantial number of pre-AR5 studies found that there is no indication of hysteresis behaviour of Arctic sea ice under decreasing temperatures following a possible overshoot of a long-term temperature target (Holland et al., 2006; Schröder and Connolley, 2007; Armour et al., 2011; Sedláček et al., 2011; Tietsche et al., 2011; Boucher et al., 2012; Ridley et al., 2012). In particular, the relationship between Arctic sea ice coverage and GMST was found to be indistinguishable between a warming scenario and a cooling scenario. These results have been confirmed by post-AR5 studies (Li et al., 2013; Jahn, 2018), which implies high confidence that an intermediate temperature overshoot has no long-term consequences for Arctic sea ice coverage.

https://www.ipcc.ch/sr15/chapter/chapter-3/
Section 3.3.8

which looks slightly different. Perhaps I was quoting a draft before

[Archimid]

that there is no indication of hysteresis behaviour of Arctic sea ice

Tiesche et al uses RCP 8.5 and he gets an ice free Arctic by 2070, yet in the very same quote, Jahn gets the attached table, RCP 8.5 is in blue. An ice free Arctic before 2050 is a virtual certainty and an ice free Arctic after 2070 is very unlikely.

Tiesche's model has a huge heat sink. This heat sink is forcing ice growth. Even with this heat sink he gets a memory. Imagine without forcing ice growth and including lateral heat transport?

[petm]

Yes, but is that the sense used by IPCC in saying ""there is no indication of hysteresis behavior of Arctic sea ice"?

They're saying that if the mean temp increase temporarily overshoots 1.5 C, there will temporarily be more frequent ice-free years, but once the temp cools back down to +1.5, the ice will re-form. I.e., No tipping point, no hysteresis.

Even if correct (doubtful), this is moot. Warming will not stop or even slow at 1.5 or 2, and will certainly not reverse. Keep in mind that the IPCC report is as much political as scientific, especially the summaries. RPC2.6 (the pathway supposedly leading to only +1.5) assumes net negative anthropogenic GHG emissions by 2070. Even if that were achievable (it isn't), very substantial positive feedbacks (e.g., methane release via permafrost thermokarst lakes) are not included in the modeling. 

[DrTskoul]

Yeah, but keep in mind the dynamics of temp rising 1.5C, are much slower (many decades ) than the ice dynamics ( a few years )...

[petm]

Agreed. Which is why this all seems so ridiculous to me. We're currently below +1C, yet are rapidly approaching ice-free events. Even if we could magically stabilize at current average temps, once the system was given time to reach equilibrium, is there any doubt that we would have regular ice-free summers, maybe every year?

Personally, this is one of the main reasons I am paying attention to the arctic ice. It's a useful source of intuition about just how far off the current models are...

[DrTskoul]

In some other thread I had mentioned it might be bouncing a while depending on weather for a few year after it reaches a critical threshold (maybe it did in 2012), but we will not have any winter ice free zones, at least not before it warms much more... the time scales between extreme weather and ice bounce back might be getting closer and commensurate (eg 2007 2012 2016 2019 ) ...

[dnem]

This question of hysteresis is one of the most important facing the global climate system.  And also somewhat moot.  In my understanding, a system exhibits hysteresis if a perturbation can push it from one state into another state from which it cannot return to the previous state.  The issue of memory is not relevant.  It's not "will it take some time to recover to the previous state?". It's CAN it return?

But all this is predicated on the variable that is doing the pushing being drawn from some fixed, known distribution. A big push may come up from time to time, but that's because it was drawn from the tails of the fixed, known distribution.  But that is not the situation we are in.  The distribution is shifting every year, with the chance of an ever greater push increasing every year.

I get that the modeling paper applied a huge push (ice free on July 1), and that it showed recovery.  But in a system with ever lower albedo, ever more accumulated heat, ever greater chances for storms and mixing, ever warmer, cloudier winters, etc., hysteresis is not necessary to end up with a state change.

[DrTskoul]

The question they tried to answer is "do such basins exist?". That is why from a simulated trajectory of a path to free ice the picked four representative states with progressively less ice in July and run those ice removal experiments starting from those states.

The assumption is that as the ice reduces due to GHG forcing adjacent dynamic basins that might be connected to the main trajectory with barriers similar to what is depicted in the previous comment. The ice removal experiments did not reveal such dynamic basins.

[Archimid]

The ice removal experiments did not reveal such dynamic basins.

Why would it? The Arctic system is perturbed once and then the climate system runs cold (2070 ice free) and eliminates the tipping point. Even in these idealized scenarios the system retains a memory that lasts many cycles. That memory combined with the increasing forcing is a tipping point all by itself.

And the most recent science makes it clear even when the IPCC doesn't notice it. There are heat sinks in these models that cause the ice free state to be delayed until 2070 while the observations and most recent models are going the opposite way.

Remove the heatsinks ( make the model be ice free before 2050) and then you will see the basins.

[DrTskoul]

The ice removal experiments did not reveal such dynamic basins.

Why would it? The Arctic system is perturbed once and then the climate system runs cold (2070 ice free) and eliminates the tipping point. Even in these idealized scenarios the system retains a memory that lasts many cycles. That memory combined with the increasing forcing is a tipping point all by itself.

And the most recent science makes it clear even when the IPCC doesn't notice it. There are heat sinks in these models that cause the ice free state to be delayed until 2070 while the observations and most recent models are going the opposite way.

Remove the heatsinks ( make the model be ice free before 2050) and then you will see the basins.

...now I know why I dont get into these discussions....

The forcing is the "artifical" part of the model. The heat sinks are the same with or without ice loss...you want free ice faster, you increase the forcing.

[Archimid]

The heat sinks are the same with or without ice loss

You haven't answered my question.

Why Tiesche et al got a "maximum perturbation" on July 1st and not June 10th? Both dates get the same insolation.  By June 1st the only heat sink left in the Arctic is the ice.

https://earth.nullschool.net/#2019/06/01/1500Z/wind/surface/level/overlay=temp/orthographic=-63.13,72.87,774


However, in Tiesche's model, when the ice is removed before July 1st the ice grows back.  This is justified by saying that July 1st must be the date of maximum albedo. That is simply not true. Just look at DMI N80. By June 1st the Arctic already reached its's max surface temperature.

Please explain to me what I'm missing.

[DrTskoul]

The heat sinks are the same with or without ice loss

You haven't answered my question.

Why Tiesche et al got a "maximum perturbation" on July 1st and not June 10th? Both dates get the same insolation.  By June 1st the only heat sink left in the Arctic is the ice.

https://earth.nullschool.net/#2019/06/01/1500Z/wind/surface/level/overlay=temp/orthographic=-63.13,72.87,774


However, in Tiesche's model, when the ice is removed before July 1st the ice grows back.  This is justified by saying that July 1st must be the date of maximum albedo. That is simply not true. Just look at DMI N80. By June 1st the Arctic already reached its's max surface temperature.

Please explain to me what I'm missing.

As the said in the paper they chose something in an arbitrary fashion.  Too early and some area will develop a thin ice skin... so july 1st seemed ok.

A single heat sink in the arctic?? I think you have your heat sinks definition mixed up. In their model there are 6 heat reservoir in the arctic ( top of atmosphere, bottom atmosphere, top ocean, deep ocean and ice ). Outside the arctic there are 3 , ocean , atmosphere, space. All are connected through convective and radiative transfer of energy and convective transfer of mass. That is the simplest description without going too deeply in the specific model they used.

[Archimid]

As the said in the paper they chose something in an arbitrary fashion.  Too early and some area will develop a thin ice skin... so july 1st seemed ok.

That is not what the paper says. The paper says:

The start date is chosen such that the effect of the perturbation is maximal: starting from ice‐free conditions earlier in the year leads to immediate re‐freezing, and hence both earlier and later start dates imply shorter exposure of open water to sunlight, and a less pronounced ice–albedo effect.

The start date was not chosen arbitrarily, at least that is not what the author claims. The start date was chosen at the point that the maximal perturbation was detected.

 We know from observations that the the northern most part of the Arctic, North of 80, maximal temperatures are reached before June 1st.

We also know from Jahn, 2018 that 2070 is at least 20 years too late, thus we know there is heat sink in Tiesche 2011 that is not accounted for.

My hope is that the reason for the July 1st discrepancy can help account for that heat sink.

In their model there are 6 heat reservoir in the arctic ( top of atmosphere, bottom atmosphere, top ocean, deep ocean and ice ).

Then one or more of the levels may be introducing a deficit of heat that causes refreeze when all the ice is removed in June.

That is the simplest description without going too deeply in the specific model they used.

Ok, but do you think that if all the sea ice is removed in June the ice will grow back even when DMI N80 indicates that the ice is holding temperatures steady right above freezing?

What elements of the system could cause the ice to refreeze?

To me the first obvious explanation is the cold water column , but it seems to me that upon receiving all that radiation the atmosphere/ocean interface would warm above freezing very quick, even if the rest of the water column takes longer to warm.

Uniquorn's animations of surface temperatures from March to mid July reinforces my case.

https://forum.arctic-sea-ice.net/index.php/topic,2591.msg216371.html#msg216371

There is no significant difference between temperatures N80 in June and July. In the peripheral area where there is no ice there is a significant difference in temperatures everywhere there is opened water. If all the ice was removed N80 would be as warm as N70.

[DrTskoul]

Do you think the results would be different if they started at June 28th ??

If I had to chose a starting date it would be at the beginning of a month...feels right... also they simulated 1980s conditions too. DMI 80N shows many years temperatures below freezing around days 150-160 ( first week of June ). The first reliable month is July. If I was one of the researchers, i would not spend too much time really finding the optimal day. I would start with the 1st of the month and I would choose may, june or july.

Also you are using DMI 80N a real observation as an argument about a model. Do you know the models equivalent DMI80N metric?

Some times there is no exchaustive rigor in the choice of initial conditions, a few simulations might be just enough.

Can you see the future? How do you know 2070 is 20 years too late? Why not 25 or 15 ?

[Archimid]

Do you think the results would be different if they started at June 28th ??

No. I worry that the model used to claim no evidence for hysteresis has a heat sink that is not properly accounted for. The most current estimates for the first ice free Arctic cited by the IPCC itself confirms this heat sink.

If I had to chose a starting date it would be at the beginning of a month.

If I had to chose a starting date for maximum perturbation I would have chosen as far back in the melting season as possible to capture maximum solar power. If I would have chosen July 1st an almost random, "good enough" date I wouldn't have called it Maximum perturbation.

Also you are using DMI 80N a real observation as an argument about a model. Do you know the models equivalent DMI80N metric?

DMI N80 represents an extremely small horizontal surface area in 2D and just the surface right above the ice in 3D. It does not represent anything above, below or south of 80N.

However, it also represent the coldest place in the Arctic. It starts taking solar energy after everything below it and the sun stops shinning there before anything to it's south. It is also adjacent to the last great ice sheet in the North Hemisphere, Greenland.  Not to mention it is sitting right over some of the deepest waters in the world.

If there is anywhere were refreezing should start in June, is there. However temperatures are at maximum in early June, in the coldest place in the Arctic. The only place where freezing takes place in June is the ice healing itself. But there is no ice to heal itself.

In the real world of today, if all the ice is magically removed in June the ice will not return until November, probably longer.

Can you see the future? How do you know 2070 is 20 years too late? Why not 25 or 15 ?

From the IPCC, according to crandles' post above:

In particular, the relationship between Arctic sea ice coverage and GMST was found to be indistinguishable between a warming scenario and a cooling scenario. These results have been confirmed by post-AR5 studies (Li et al., 2013; Jahn, 2018), which implies high confidence that an intermediate temperature overshoot has no long-term consequences for Arctic sea ice coverage.

I checked Jahn 2018 because it was the most recent and I found the attached graph. RCP 8.5 leads to an ice free Arctic much sooner than Tiesche et al predicts. This is proof that there are heat sinks in Tiesche. The models used to claim no hysteresis are running cold with respect to the Arctic ice.

If there was a threshold behavior change the heatsinks in the models are likely hiding them, like they hid June warmth.

Even with the heatsinks we get several years of memory.

[crandles]

The ice removal experiments did not reveal such dynamic basins.

Why would it? The Arctic system is perturbed once and then the climate system runs cold (2070 ice free) and eliminates the tipping point. Even in these idealized scenarios the system retains a memory that lasts many cycles. That memory combined with the increasing forcing is a tipping point all by itself.

And the most recent science makes it clear even when the IPCC doesn't notice it. There are heat sinks in these models that cause the ice free state to be delayed until 2070 while the observations and most recent models are going the opposite way.

Remove the heatsinks ( make the model be ice free before 2050) and then you will see the basins.

The speculation is that the albedo effect of ice can lead to two or more stable states. Without sea ice cover, albedo is low possibly supporting a low or zero level of ice. With the sea ice, albedo is high so heat is reflected and a higher level of ice might be supported at the same GHG level. This suggests that the perturbation that should be applied is removing the ice rather than some random perturbation that may well be unsuited to testing for different basins.

With a really large perturbation the effects are declining over up to 8 years. Where it is as many as 8, this is because the perturbation has affected ocean temperatures below 50m depth. The sort of perturbation that could occur from some bizarre weather are much smaller and things tend to return to the model norm within 2 years if not by a year later. This is not much delay at all from an air temperature perturbation.

>[Inappropriate] heat sinks in these models
That is just a conclusion you want to draw. Have you emailed the corresponding author, Tietsche to find out what was done to support that wording?

The GC models are all over the place with quantity of ice. The model Schroeder & Connolley used had far too little ice such that they had to use preindustial CO2 levels to get sensible level of ice. Tietsche's model has ice til 2070,  Jahn goes ice free much sooner. If the models are all over the place, it is a bit silly to say I prefer model used by Jahn and use that as date when models say Arctic will be seasonally ice free. If all the models agree that there is strong deceleration in rate of ice loss before frequent seasonally ice free conditions are reached. Then this is something that has backing from lots of models and is likely true.

>Remove the heatsinks ( make the model be ice free before 2050) and then you will see the basins.

Is this based on WintonI. Eisenman and J. S. Wettlaufer 2009 (linked below)? or some other paper or is it just an assertion you think likely or want to believe?

[DrTskoul]

A very interesting theoretical analysis to help:

Nonlinear threshold behavior during the loss of Arctic sea ice
https://doi.org/10.1073/pnas.0806887106

Conclusions: Our analysis suggests that a sea-ice bifurcation threshold (or “tipping point”) caused by the ice–albedo feedback is not expected to occur in the transition from current perennial sea ice conditions to a seasonally ice-free Arctic Ocean, but that a bifurcation threshold associated with the sudden loss of the remaining seasonal ice cover may occur in response to further heating. These results may be interpreted by viewing the state of the Arctic Ocean as comprising a full seasonal cycle, which can include ice-covered periods as well as ice-free periods. The ice–albedo feedback promotes the existence of multiple states, allowing the possibility of abrupt transitions in the sea-ice cover as the Arctic is gradually forced to warm. Because a similar amount of solar radiation is incident at the surface during the first months to become ice free in a warming climate as during the final months to lose their ice in a further warmed climate, the ice–albedo feedback is similarly strong during both transitions. The asymmetry between these two transitions is associated with the fundamental nonlinearities of sea-ice thermodynamic effects, which make the Arctic climate more stable when sea ice is present than when the open ocean is exposed. Hence, when sea ice covers the Arctic Ocean during fewer months of the year, the state of the Arctic becomes less stable and more susceptible to destabilization by the ice–albedo feedback. In a warming climate, as discussed above, this causes irreversible threshold behavior during the potential distant loss of winter ice, but not during the more imminent possible loss of summer (September) ice.

The relevance of any basic theory to the actual future evolution of the complex climate system must be carefully qualified. Because the time scale associated with the sea-ice response to a change in forcing may be decadal, and the time scale associated with increasing greenhouse gas concentrations may be similar, the system may not be operating close to a steady-state. In the gradual approach to steady-state under a continual change in forcing, the difference between a region of the steady-state solution with increased sensitivity to the forcing and an actual discontinuous bifurcation threshold (as in Fig. 3) could be difficult to discern. If greenhouse gas concentrations were reduced after crossing a bifurcation threshold, however, the possible irreversibility of the trajectory would certainly be expected to be relevant.

The model questions are given and easy to follow

Simply instability only manifests itself when we get closer to a winter ice-free state. Ice thermodynamics are the stabilizing force of the ice-albedo feedback. In a background climatological state similar to the experiments of Tietsche, removing all the ice in the summer did not lead to instability and bifurcation.

[Archimid]

The speculation is that the albedo effect of ice can lead to two or more stable states.

It's not just Albedo.

What causes the tipping point, threshold behavior or rate independent hysteresis  is:

A. Summer

 1. Albedo
 2. Latent heat of fusion
 3. Increased heat intrusions from the breakdown of typical the atmospheric currents.
 4. Ice volume ends at progressively lower state

B. Winter

 1. The ice starts at progressively lower state
 2. Ice formation starts late.
 3. Ice area growth will be ridiculously fast but it will thicken slow because of the warmer air, warmer water 
 4. Increased heat intrusions from the hemisphere

Back to A 1

The following year the ice will be thin, weak and warm, quickly melted, increasing the temperatures everywhere.

[Archimid]

Ok. Please reconcile for me Figure 1. B of I. Eisenman and J. S. Wettlaufer linked above with Tiesche et al July 1st explanation:

The start date is chosen such that the effect of the perturbation is maximal: starting from ice‐free conditions earlier in the year leads to immediate re‐freezing, and hence both earlier and later start dates imply shorter exposure of open water to sunlight, and a less pronounced ice–albedo effect.

To me they are in disagreement. Tiesche et al is significantly biased towards cold

Simply instability only manifests itself when we get closer to a winter ice-free state.

Read again.

Our analysis suggests that a sea-ice bifurcation threshold (or “tipping point”) caused by the ice–albedo feedback is not expected to occur in the transition from current perennial sea ice conditions to a seasonally ice-free Arctic Ocean, but that a bifurcation threshold associated with the sudden loss of the remaining seasonal ice cover may occur in response to further heating.

Ice Albedo by itself is not enough to cause a bifurcation in this model, but further warming is. Ice-albedo then reinforces that later warming.

Ice thermodynamics are the stabilizing force of the ice-albedo feedback

and global warming the destabilizing force.

In a background climatological state similar to the experiments of Tietsche, removing all the ice in the summer did not lead to instability and bifurcation.

The climatological state of Tiesche have proven much colder than reality. July 1st is not maximum perturbation in no way shape or form in the real world, only in a old model that predicts ice free summers by 2070.
 

[DrTskoul]

Ok, i provide literature to back their argument, and you provide your own thoughts. I suggest that you have enough energy and thoughts then to write a new model, run experimenta and publish the result as a rebuttal since you absolutely know what's wrong and what's right. The scientific community will thank you for your contributions....handwaving does not cut it any more...

[Archimid]

I don't do personal, so forget about emails.

I've already done very simple linear models and posted them in ASIF.

I'm not a specialist, I'm a jack of all trades, master of none. That makes me excellent at knowing what is wrong, but not what is right.

I'm reviewing the literature using the literature. Isn't that what the science is about. Hopefully other in the position to write this stuff up gain something from my writings.

[DrTskoul]

How do your simple linear models compare to:

https://doi.org/10.1073/pnas.0806887106

I am just leery of people that they "know" what is wrong and have a simple model that proves it , in the work of people that spent their careers developing..

[philopek]

How do your simple linear models compare to:

https://doi.org/10.1073/pnas.0806887106

I am just leery of people that they "know" what is wrong and have a simple model that proves it , in the work of people that spent their careers developing..

Avoiding mistakes by knowing what's wrong would safe 90% of all resources wasted on planet earth while during the time needed to get things right we could get lost in space.

Many things work after the "negative principle" means by falsification and then concentrate on falsifying the rest until very little remains to eventually find out what's right.

Hence I'm leery of poeple who think that because they need decades to understand something that others cannot understand way more in a split second because they know most of what's certainly wrong.

To know all kind of theories is certainly a very safe approach for many things. I have a friend who learned to play chess, reading books, studying openings and end-games over years while I have almost no clue about those things but play spontaneously.

After keeping a statistic over 50 years of playing together i win more by 10%. He starts perfect
and if he has an advantage towards the end his knowledge of endgames make me lose, while my I win games in the middle where intuition and imagination produce new situation that are non-standard and are mostly not stored in the brains of the theoreticians.

What I'm trying to convey, both are needed and only because one tries to see himself above the other does it take so long for humankind to achieve goals in an efficient manner.

Those who are knowledgeable by learning (working) are often envious of those who apparently get things out of their fingertips with apparent ease and those who have perhaps the gift to think faster and in larger contexts with a super-memory, lose patience too often and consider the workers (those who know by spending time to understand) as less intelligent.

Both are wrong but to be leery about each other is building a huge borderline with such high obstacles that are almost impossible to overcome.

[Archimid]

How do your simple linear models compare to:

It is difficult to compare my simple linear models to full blown climate models. Mine is basic arithmetic sprinkled with basic excel sheets. The models used in the paper you linked, like Tiesche and all other authors, are marvels of modern applied mathematics and physics. I mean, they are trying to literally model the world .

Despite being marvels, they are models, thus very likely wrong. Finding out how they are wrong makes them more useful. Trusting them blindly makes them wrong.

I am just leery of people that they "know" what is wrong and have a simple model that proves it

I know nothing, relative to all there is to know, but I do know some things.

I don't have a simple model that proves anything nor have I ever claim to have one.

However, my simple model helps me understand some of the most important elements of the Arctic system.  What I see in the volume trends and the rest the climate and arctic reinforce my belief that the Arctic is heading to rapid collapse. The models say no rapid collapse. What am I missing? I'm trying to find out.

Then I start reading the sources for the IPCC and they cite pre industrial climates
, a model that ignores the June sun and a model that does not remove the ice at all.

Despite that all three models, even during cold pre-industrial times, exhibit memory. But the IPCC says there is no sign of hysteresis.