When Will We See <1,000 km^2 Arctic Sea Ice Extent

[jai mitchell]

To get a sense of the expected future forcing rates of sea ice albedo feedbacks under 2XCO2 scenarios.  Please pick which you think is most likely.  By selecting on of the Sept. Minimum choices you are saying that you do not believe a June 21 ice free condition is likely this century.

leave a comment if you want another option or have a different scenario.

 

[oren]

A. I would choose ice-free Sept before 2030. 2020 is too early for my taste. 2040 too late.
B. Personally I would not mix together the two issues in the same poll, ice-free Sept in a decade vs. ice-free June 21st this century.

[jai mitchell]

I am of the opinion that if we don't see September ice free by 2025 or so we probably won't see June 21 ice free this decade.  though It is very murky at late decades.  To date I am the only person I have read who has considered ice free arctic conditions on June 21st.  to me this is very significant as the ice-free date moves further and further toward the solstice the amount of increased absorbed energy grows significantly.

[Tor Bejnar]

My guess would be a functionally sea-ice-free day in the Arctic by September or October 2025.  Influenced by Chris R., I don't expect such a day within the next five years.  My "forecasting" ability is so bad, however, that I'm prepared for Neven's "not this year" to fail.   

I too think you should have more choices and not mix the concepts.

[slow wing]

Do you really mean just 1000 km^2?
Isn't the conventional definition of 'functionally ice-free' set at  1,000,000 km^2?

[epiphyte]

Could we have an option for "I know enough to know I don't know." ?

[Michael Hauber]

IPCC projections are for an ice free Arctic in September roughly near 2050.  However this is under rcp 8.5, which I consider quite unlikely.  Rcp 8.5 requires that the current rapid growth of renewables never gets high enough to make a difference.  It also requires that climate change never has an economic impact significant enough to noticeably slow 'Business as Usual'.   This paper contains projections of arctic for RCP 4.5 - with 9 out of 14 models becoming ice free by 2069, and 3 of the models never becoming ice free.  There is also a statement that the Arctic becomes ice free in September at 2.4 deg of warming (with a range from as low as 0.4 up to 6.2), and in March at 8.2.  If I split the difference I get 5.3 degree of warming for ice free in June.  Even RCP 8.5 does not get to about 5 degree of warming around 2150. Although the upper end of RCP 8.5 does get there around 2100. 

Another way to estimate June ice free is by noting that we are currently about 30 years behind September for historical trends.  For RCP 8.5 I would expect this difference to be roughly maintained, so June ice free could be reached for 8.5 around 2080.  But under 4.5 or 6 I'd expect the slow down in forcing to increase the 30 year gap significantly.

As a side point 560ppm is achieved in rcp 8.5 about 2050, but is never reached under rcp4.5, and just barely reached in rcp 6.0.  Unless we follow 8.5 there is an extremely wide possible range for when we get there.

Finally there is the point of ice free once vs ice free consistently.  I believe the model projections are a reasonable prediction of when we might see consistent ice free conditions.  I suspect rcp 4.5 is a little optimistic and that rcp 6 is more reasonable, but could only find the rcp 4.5 projection.  I'll guess rcp 6 may see ice free somewhere between 2050 and 2075.  However year to year variability could see at least one nearly ice free year (<1,000,000 sq km extent) between 2020 and 2040.  I do have a suspicion that perfect melt conditions could see nearly ice free any year from now (and perhaps back as far as 2008), but variability observed to date doesn't seem large enough.  There is the possibility that variability grows as ice gets lower....

[jai mitchell]

Another way to estimate June ice free is by noting that we are currently about 30 years behind September for historical trends.

what?

[Carex]

Three years ago I put my money down for 2020.  So, I'm stuck with that.  I think MH's final paragraph is critical.  A weather driven ice free September is now possible in any given year.  A climate driven ice free September is not likely until 2030 to 2050, and I would trend toward the later part of that period.  As climate warms weather should become less predictable, a somewhat balanced system becomes disturbed and more chaotic until a new equilibrium is reached, which will be a long time coming.   Nice to have a topic that we inexpert readers can contribute to, even if our ideas turn out to be BS.  Ahh, the wonders of prognostication.

[ChrisReynolds]

Thanks for the poll Jai.

I've voted for before 2040, 1000km^2 is not my preferred level, I'd be more confident with 1 million km^2. But there is a good chance of very low extents in the 2030s, and as Michael Hauber points out - weather may just tip it.

BTW - as April peak thickness decreases a given thinning in the spring/summer will be able to reveal more open water. But equally under the same thinning, a poor melt weather year will see a wild upswing in extent. So far only 2007 and 2009 after and 2012 and 2013 add to the mathematical indications of variability increasing. By the mid 2020s we will see much more evidence of an increase in variability.

There are some exciting years ahead.

[Michael Hauber]

Another way to estimate June ice free is by noting that we are currently about 30 years behind September for historical trends.

what?

I meant that June this year is roughly the same as September 30 years ago.  Looking at Neven's long term chart section I'm not sure whether I misread the area and extent chart or was looking at the volume chart.

[Polynya88]

Why no option for "not in this millennium"? RE: This paper contains projections of arctic for RCP 4.5 - with 9 out of 14 models becoming ice free by 2069, and 3 of the models never becoming ice free. 40 years ago the ice in the Arctic was thick, extensive, and no one saw it improving; it was stated that regular commercial arctic shipping was, and would remain,  a non-starter. (Beaufort Sea offshore oil exploration summary - 1974) Thus to now suggest we are (again?) confident of weather patterns/sea ice conditions 40 years into the future seems a bit naïve. It's not the unknowns that will foil our forecasts, it is the unknown unknowns...

[DavidR]

I  haven't voted in this because I really don't understand the question. 
In terms of likelihood of occurring I would put options 1,2,4 as about a 99% probability.
Option 3 is about a 50% probability and Option 5 about a 25% probability.

Option 5 is impossible to assess because we really  don't know how refreeze will be affected by three factors.

1. An ice free North pole in September, (within 5 years)
2. An ice free arctic in September, (within 10 years)
2. Winter extent not extending out to the Bering. 

The decline in the winter maximum has been fairly steady at -0.04 M km^2 for the past 35 years however this rate of decline may change significantly once these factors come in to play. This rate of decline would have to increase significantly for an ice free Arctic in June 2050 to become a reality.

I suspect the current El Nino will, if it continues until early 2016, lead to an ice free Arctic by  2020. The exceptional thing about 2015 so far is that it hasn't been exceptionally  warm in the Arctic (about 8th warmest) and yet we are still at record low levels of extent.

[kiwichick16]

hi guys, first post here.  Kia ora.

I voted for icefree by 2020; because I like a little flutter

and because I don't think this "pause" will last as long as the previous 2 (1880- 1910  and 1940 - 1970 )

[kiwichick16]

and I've just read the latest ENSO report from the Australian BOM   

[jai mitchell]

Why no option for "not in this millennium"? RE: This paper contains projections of arctic for RCP 4.5 - with 9 out of 14 models becoming ice free by 2069, and 3 of the models never becoming ice free. 40 years ago the ice in the Arctic was thick, extensive, and no one saw it improving; it was stated that regular commercial arctic shipping was, and would remain,  a non-starter. (Beaufort Sea offshore oil exploration summary - 1974) Thus to now suggest we are (again?) confident of weather patterns/sea ice conditions 40 years into the future seems a bit naïve. It's not the unknowns that will foil our forecasts, it is the unknown unknowns...

Because, "aliens will arrive and fix our global energy imbalance problem" didn't seem like a realistic answer???

[jai mitchell]

I  haven't voted in this because I really don't understand the question. 
In terms of likelihood of occurring I would put options 1,2,4 as about a 99% probability.
Option 3 is about a 50% probability and Option 5 about a 25% probability.

Option 5 is impossible to assess because we really  don't know how refreeze will be affected by three factors.

1. An ice free North pole in September, (within 5 years)
2. An ice free arctic in September, (within 10 years)
2. Winter extent not extending out to the Bering. 

The decline in the winter maximum has been fairly steady at -0.04 M km^2 for the past 35 years however this rate of decline may change significantly once these factors come in to play. This rate of decline would have to increase significantly for an ice free Arctic in June 2050 to become a reality.

I suspect the current El Nino will, if it continues until early 2016, lead to an ice free Arctic by  2020. The exceptional thing about 2015 so far is that it hasn't been exceptionally  warm in the Arctic (about 8th warmest) and yet we are still at record low levels of extent.

excellent analysis david.  I also agree with your timeline.  However, the fact that our current Top of Atmosphere energy imbalance is significantly inhibited by anthropogenic sufate emissions AND the 10-year time lag that slows down the water vapor and lapse rate feedbacks means that the current shock to the arctic that will occur this year (and the next few) due to china's attack on smog coupled with the growing el nino will produce a step-change in climate forcing induced weather patterns that will destroy the arctic ice in 3-7 years and show us that we severely underestimated the impact of June snow cover anomalies on arctic sea ice.  I am very certain that we will see July 15th ice free conditions by 2025.

[Buddy]

THAT......will be a LOT of dark blue ocean to absorb a LOT of heat that would otherwise be reflected.

[DavidR]

excellent analysis david.  I also agree with your timeline.  However, the fact that our current Top of Atmosphere energy imbalance is significantly inhibited by anthropogenic sufate emissions AND the 10-year time lag that slows down the water vapor and lapse rate feedbacks means that the current shock to the arctic that will occur this year (and the next few) due to china's attack on smog coupled with the growing el nino will produce a step-change in climate forcing induced weather patterns that will destroy the arctic ice in 3-7 years and show us that we severely underestimated the impact of June snow cover anomalies on arctic sea ice.  I am very certain that we will see July 15th ice free conditions by 2025.

You  may  be right about 2025, once there is open water between the ice and the North Pole, refreeze should start to slow as the ice has to extend to the pole before starting to freeze on the siberian side of the arctic.   I  am assuming that refreeze occurs most  easily  where there is existing ice or land to extend from.  By the time we have an ice free arctic there would be no significant ice above about 85N so that delay would be significant.  Its not  so much that thye winter extent  might not be close to  current expected levels but that the refreeze would be more fragile and prone to melt in the next year.  We would also  see a much more fragmented and mobile area of retained ice that  would move more easily into  areas of higher melt in peripheral seas.

[Siffy]

IPCC projections for an Ice free Arctic are complete and utter rubbish. Let's be completely honest here their projections are moronically behind observational evidence and using the IPCCs projections for the state of the Arctic as evidence of what to expect in thirty years is incredibly silly when observational data show cases current conditions to be worse than what is just projected 20 years later.



*shrug*

[crandles]

By selecting on of the Sept. Minimum choices you are saying that you do not believe a June 21 ice free condition is likely this century.

When do you expect 560ppmv CO2 to be reached?

Installation of renewables may well slow growth of CO2 to less than 2ppmv per year so 160 extra ppmv may take over 80 years ie past 2095, add the 10 years takes us to 2105.

I don't see why someone couldn't believe September ice free by ~2025 and June 21 ice free by ~2105 or a bit later if CO2 levels rise a bit slower due to renewables installation.  In this case they could answer both Sept by 2040 and June 21 by 10 years after reaching 560ppmv

[Jim Pettit]

Why no option for "not in this millennium"? RE: This paper contains projections of arctic for RCP 4.5 - with 9 out of 14 models becoming ice free by 2069, and 3 of the models never becoming ice free. 40 years ago the ice in the Arctic was thick, extensive, and no one saw it improving; it was stated that regular commercial arctic shipping was, and would remain,  a non-starter. (Beaufort Sea offshore oil exploration summary - 1974) Thus to now suggest we are (again?) confident of weather patterns/sea ice conditions 40 years into the future seems a bit naïve. It's not the unknowns that will foil our forecasts, it is the unknown unknowns...

Are you implying that because early, very conservative forecasts made with a minimum of data points severely underestimated the rate at which the NH sea ice would disappear, we should assume that modern forecasts--still conservative, but created with much more powerful computers crunching vastly more data based on several extra decades of observation--are overestimating the rate of sea ice decline?

I don't think so.

Arctic sea ice is a goner. That doesn't mean we're going to soon experience tropical weather at the poles (ala the moronic Waterworld). Neither does it mean that the months-long nights at the poles will no longer bring deep freezes or patches of ultra-thick ice. But the long-term downward spiral continues; summer ice will disappear, then more and more spring and autumn ice will follow suit. Short of jai's energy imbalance-fixing aliens, it's a certainty...

[jai mitchell]

By selecting on of the Sept. Minimum choices you are saying that you do not believe a June 21 ice free condition is likely this century.

When do you expect 560ppmv CO2 to be reached?

Installation of renewables may well slow growth of CO2 to less than 2ppmv per year so 160 extra ppmv may take over 80 years ie past 2095, add the 10 years takes us to 2105.

I don't see why someone couldn't believe September ice free by ~2025 and June 21 ice free by ~2105 or a bit later if CO2 levels rise a bit slower due to renewables installation.  In this case they could answer both Sept by 2040 and June 21 by 10 years after reaching 560ppmv

The purpose of this is to generate some kind of constraint on the total arctic sea ice forcing feedback that can be attributed to ECS 2XCO2.  Without projecting out so far that you may as well be talking about ESS.

My cognition of the current rapid increasing rates of top of atmosphere imbalance being driven by catching up with the last 10 years of CO2equivalent emissions AND the rapid decrease in SO2 emissions from China shows that we will experience a <1.0 Mkm^2 sea ice extent before 2020.

under this scenario, a June 21 arctic ice free state is assured some time around 2065.

(thanks for the units check, Crandals)

---------

Cross post the calcs over at Then There's Physics:  https://andthentheresphysics.wordpress.com/2015/05/19/lukewarmers-a-follow-up/#comment-56713

“ECS is actually 6C when arctic sea ice feedbacks are taken into account”

you are sorely mistaken if you think that we wont experience <1,000,000 km^2 of arctic sea ice by June 21 under 2XCO2 forcing. This means an additional globally averaged albedo feedback of about 0.23 watts per meter squared for each 1,000,000 km^2 loss. (may be as high as 0.3 but I low-estimate) The 1970-1980 mean of ice for June 21 is about 11,000 KM^2 so this ice free state is equivalent to about 2.3 watts per meter squared globally averaged forcing feedbacks compared to pre-industrial.

Add that to the 3.7 of 2XCO2 and you get 6.0 Watts per meter squared without other feedbacks This calculation neglects the reduction of oceanic Dimethyl Sulfide production which is projected to yield (median) +0.3 C of warming feedback, as well as other feedbacks such as carbon cycle, frozen soils and far-infrared emissivity reductions and increased arctic algae blooms causing further albedo effects.

not to mention northern hemisphere snow-cover anomalies.

since our best estimate so far is 0.6C of globally averaged warming for each 1.0 Watt per meter squared that is 3.6C of warming just for CO2, WV/Lapse Rate and Arctic sea ice. Add the 0.3 DMS feedback, emissivity and algae bloom effects and you end up with ~4.3C as a baseline response.

Then you can add cloud feedbacks, snow cover albedo effects, frozen soil effects and carbon-cycle effects.

as you can see the 6.0C ECS is a low-end estimate. This is why we will have to engage in geoengineering activities, including dimming and CO2 atmospheric extraction.

------
Crandals,

I believe that we will cross 2XCO2 sometime around 2080, this will be due to carbon cycle feedbacks as the human population will only be about 200 million people by then.  (unless. . .unless. . .

[crandles]

My cognition of the current rapid increasing rates of top of atmosphere imbalance being driven by catching up with the last 10 years of CO2equivalent emissions AND the rapid decrease in SO2 emissions from China shows that we will experience a <1.0 Mkm^2 sea ice extent before 2020.

under this scenario, a June 21 arctic ice free state is assured some time around 2065.

Do you have some data showing the importance of decrease in SO2 emissions from China?

Arctic zonal mean Stratospheric Aerosol Optical Thickness at
http://data.giss.nasa.gov/modelforce/strataer/tau.map_2012.12.txt
is showing zeros for most years since 2000 with some 0.0001 numbers for Oct 2008 and second half of 2009 and 2011. Doesn't look like much scope for reduction there.

Is it tropospheric SO2 that is more of a problem and does enough of it reach Arctic latitudes? If enough of it does reach Arctic latitudes, does it reduce melting point of ice so reduction in pollution would make ice harder to melt? Is this negligible compared to reduction in insolation? If it is negligible compared to insolation, might wild fire smoke take over and become more important in keeping insolation down?

[jai mitchell]

Do you have some data showing the importance of decrease in SO2 emissions from China?

indirect evidence derived from http://onlinelibrary.wiley.com/doi/10.1002/2014GL062015/abstract


Also: 

Is it tropospheric SO2 that is more of a problem and does enough of it reach Arctic latitudes? If enough of it does reach Arctic latitudes, does it reduce melting point of ice so reduction in pollution would make ice harder to melt? Is this negligible compared to reduction in insolation? If it is negligible compared to insolation, might wild fire smoke take over and become more important in keeping insolation down?

yes, and yes, apparently, it affects incoming solar insolation signfiicantly, increasing low level haze reducing incoming insolation but producing warmer winters as well, see reference above.
As far as I have seen, the wildfires are a null effect since the increased black carbon is a significant forcing while the aerosol haze a negative forcing.

The historical record is clear.  The effect of the aerosol haze on the El Nino Southern Oscillation as well as the arctic feedback response is resoundingly clear.

[ktonine]

The historical record is clear.  The effect of the aerosol haze on the El Nino Southern Oscillation as well as the arctic feedback response is resoundingly clear.

I don't think it's clear at all. For instance, Global atmospheric sulfur budget under volcanically quiescent conditions: Aerosol-chemistry-climate model predictions and validation
"...sensitivity simulations reveal that anticipated increases in anthropogenic SO2 emissions in China and India have a larger influence on stratospheric aerosols than the same increase in Western Europe or the U.S., due to deep convection in the western Pacific region. However, even a doubling of Chinese and Indian emissions is predicted to increase the stratospheric background aerosol burden only by 9%. In contrast, small to moderate volcanic eruptions, such as that of Nabro in 2011, may easily double the stratospheric aerosol loading."

And if you look at a graph of global SO2 emissions there appears to be little correlation with global temperature.  Putting an arrow on a graph that says "Clean Air Regulations" hardly suffices as evidence. 

[jai mitchell]

The main impact are tropospheric, not stratospheric.  however, even the stratospheric backscatter has been on the increase through 2007 as shown here  http://onlinelibrary.wiley.com/doi/10.1029/2009GL039008/abstract

Since about 2000, an increase of 4–7% per year in the aerosol backscatter in the altitude range 20–30 km has been detected at both Mauna Loa and Boulder.

Looking at pure emissions data is not correct, the government statistics from china underrepresent the emissions and the lack of emissions controls are still rampant though the country.  Their coal and liquid fuel consumption was increasing at a 5% annual rate 2005-2014

This is more recent http://iopscience.iop.org/1748-9326/8/1/014003

Still underestimates some what but at least is more recent.

ktonine,

you said,

"And if you look at a graph of global SO2 emissions there appears to be little correlation with global temperature."

I do not believe that you actually have.

[Michael Hauber]

IPCC projections for an Ice free Arctic are complete and utter rubbish. Let's be completely honest here their projections are moronically behind observational evidence and using the IPCCs projections for the state of the Arctic as evidence of what to expect in thirty years is incredibly silly when observational data show cases current conditions to be worse than what is just projected 20 years later.



*shrug*

Those comparisons are out of date.  More recent comparisons of model vs observations looks like:



source

Also when I chased down the source of that comparison, it comes from Real Climate.  It is not a chart of IPCC models in general, but from one specific model being CCSM4 (that was used in AR4).

[Michael Hauber]

The main impact are tropospheric, not stratospheric.  however, even the stratospheric backscatter has been on the increase through 2007 as shown here  http://onlinelibrary.wiley.com/doi/10.1029/2009GL039008/abstract

Since about 2000, an increase of 4–7% per year in the aerosol backscatter in the altitude range 20–30 km has been detected at both Mauna Loa and Boulder.

The author of this paper rules out volcanic contribution using the logic that no large volcanic eruptions have occurred.  However several researchers eg Bruhl have investigated this data and find that the increase is more likely to have been caused by volcanos than by anthropogenic emissions.

Looking at pure emissions data is not correct, the government statistics from china underrepresent the emissions and the lack of emissions controls are still rampant though the country.  Their coal and liquid fuel consumption was increasing at a 5% annual rate 2005-2014

This is more recent http://iopscience.iop.org/1748-9326/8/1/014003

Still underestimates some what but at least is more recent.

From that paper:

After increasing until about 2006, we estimate a declining trend continuing until 2011

Interesting.  The hiatus really had its start around 2008, with the years from 2002 to 2006 all being quite warm.  Maybe a lag or natural variation can reconcile these, however if the data is correct it seems to me that the aersol effect is not strong enough to have a consistent impact on global temperatures within a five year time frame.

[Michael Hauber]

And in other news, there is some speculation that the AMO may be switching from its positive to negative phase. 

Will this matter for the Arctic?  A bit of google searching on Arctic and AMO uncovers some rubbish, but also finds an interesting article Day et al.  A key issue with the AMO is that it is basically a measure of the North Atlantic temperature with the trend removed.  So anything that impacts global temperature will drive AMO, so correlations between AMO and global temperature and related factors such as arctic ice do not provide evidence for an AMO influence as the correlation could be simply due to global temps driving both AMO and Arctic etc (thats the rubbish I referred to earlier).  However this study looks at behaviour in climate models and then compares with observed data to conclude that the AMO may have caused as much as 30% of the recent decrease in sea ice.  Or as little as 5%.

From the related article tab for that article I then find this which investigates the long wave feedback.  This paper finds that later in the century this feedback will reach a limit as atmosphere becomes closer to a black body, and that Arctic warming will slow.  A point always to watch out for when a study claims faster or slower warming or other impacts is to ask 'how much'.  On a first read of the document I can't find the answer so the answer could well be 'not enough to be worth mentioning'.

[ktonine]

ktonine,
you said,

"And if you look at a graph of global SO2 emissions there appears to be little correlation with global temperature."

I do not believe that you actually have.

Jai - if you'd followed the link I provided you'd see the figure I linked to is nearly the same as the one you provided.  Nor do I see any strong correlation - just as I stated earlier.

And if you look at a graph of global SO2 emissions there appears to be little correlation with global temperature. 

[jai mitchell]

there is a clear negative correlation to to temperature in the northern hemisphere.  Recall also that the CO2 forcing is steadily increasing during this time and there is a 10 year delay in maximum warming from that forcing effect. 

The negative correlation is clear, as SO2 started to increase temperatures stalled and then declined.  This is even more obvious in the arctic (as shown above)  As SO2 emissions stalled and then were in decrease, temperatures began to climb (again much worse in the arctic).  The link to tropospheric SO2 emissions is very clear.

[Ned W]

globally averaged albedo feedback of about 0.23 watts per meter squared for each 1,000,000 km^2 loss

How do you calculate that 0.23?  What value did you use for a top-of-atmosphere difference in albedo between ice and ocean?  Is that 0.23 *annually* averaged as well as *globally* averaged, or just for the seasonal peak?  Is the 1M km2 loss of ice area annually averaged? 

So many questions ... because this is complicated.

Hudson 2011 (via Tamino here) suggests a TOA albedo difference of 0.2 between sea ice and ocean.

Hudson 2011 says "Results show that the globally and annually averaged radiative forcing caused by the observed loss of sea ice in the Arctic between 1979 and 2007 is approximately 0.1 W/m2" and by my calculation, annually averaged Arctic sea ice area during that period declined by about 1.6 million km2. 

Tamino repeats the albedo calculation and finds 0.13 W/m2 change from 1979-2011, during which time annually averaged Arctic ice area declined by about 1.9 million km2.

Divide each change in albedo by each change in area, and you get a globally and annually averaged radiative forcing from loss of Arctic sea ice of 0.068 per million km2 (Tamino) or 0.062 per million km2 (Hudson 2011). 

Your estimate of 0.23 (which you describe as a "low estimate" is actually more than 3 times larger than these.  So I'm curious how you calculated it.

[S.Pansa]

My cognition of the current rapid increasing rates of top of atmosphere imbalance being driven by catching up with the last 10 years of CO2equivalent emissions AND the rapid decrease in SO2 emissions from China shows that we will experience a <1.0 Mkm^2 sea ice extent before 2020.

under this scenario, a June 21 arctic ice free state is assured some time around 2065.

Do you have some data showing the importance of decrease in SO2 emissions from China?

Arctic zonal mean Stratospheric Aerosol Optical Thickness at
http://data.giss.nasa.gov/modelforce/strataer/tau.map_2012.12.txt
is showing zeros for most years since 2000 with some 0.0001 numbers for Oct 2008 and second half of 2009 and 2011. Doesn't look like much scope for reduction there.

Is it tropospheric SO2 that is more of a problem and does enough of it reach Arctic latitudes? If enough of it does reach Arctic latitudes, does it reduce melting point of ice so reduction in pollution would make ice harder to melt? Is this negligible compared to reduction in insolation? If it is negligible compared to insolation, might wild fire smoke take over and become more important in keeping insolation down?

Although the following paper is not only about the influence of Chinese aerosols,  it might be of interest here, as it tries to put a number on the magnitude of the aerosol-cooling-effect in the Arctic. And all in all it is quite substantial.
From the  3°C temp rise GHGs should have caused in the Arctic between 1913 and 2012, Aerosols have offset 1.8 °C. See the article and the Figure attached below on The Carbon Brief.

The paper is:
Najafi, Attribution of Arctic temperature change to greenhouse-gas and aerosol influences, Nature 2015 (free copy here)

Abstract: The Arctic has warmed significantly more than global mean surface air temperature over recent decades1, as expected from amplification mechanisms2, 3. Previous studies have attributed the observed Arctic warming to the combined effect of greenhouse gases and other anthropogenic influences4. However, given the sensitivity of the Arctic to external forcing and the intense interest in the effects of aerosols on its climate5, 6, it is important to examine and quantify the effects of individual groups of anthropogenic forcing agents. Here we quantify the separate contributions to observed Arctic land temperature change from greenhouse gases, other anthropogenic forcing agents (which are dominated by aerosols) and natural forcing agents. We show that although increases in greenhouse-gas concentrations have driven the observed warming over the past century, approximately 60% of the greenhouse-gas-induced warming has been offset by the combined response to other anthropogenic forcings, which is substantially greater than the fraction of global greenhouse-gas-induced warming that has been offset by these forcings7, 8. The climate models considered on average simulate the amplitude of response to anthropogenic forcings well, increasing confidence in their projections of profound future Arctic climate change.

 

[jai mitchell]

0.23 comes from http://journals.ametsoc.org/doi/full/10.1175/JCLI-D-14-00042.1 which references the hudson paper you cite. 

Results obtained here indicate that in this configuration of CESM (CAM4 coupled to a slab ocean and the dynamic–thermodynamic sea ice model CICE4), approximately 3 × 1012 m2 of sea ice is lost for each kelvin of global mean warming and approximately 0.1 W m−2 of “sea ice radiative forcing” is produced by each 1012 m2 of sea ice loss,

in direct discussion with Dr. Caldeira about his value he agreed that the application of the 0.1 watts per meter squared was weighted toward september values and that if these values were instead applied to June 21 ice free conditions the increased absorbed insolation would be between 0.2 and 0.3 Watts per meter squared per million km^2 lost sea ice.

in addition, Cao et al http://journals.ametsoc.org/doi/abs/10.1175/JCLI-D-14-00389.1  Showed that the cloud impacts of hudson were significantly overstated.

[jai mitchell]

Jai - if you'd followed the link I provided you'd see the figure I linked to is nearly the same as the one you provided. 

I know, we used the same source data to generate our graphs! 

[ChrisReynolds]

S Pansa,

Thanks for that paper. And thanks to everyone else for the discussion.

[jdallen]

S Pansa,

Thanks for that paper. And thanks to everyone else for the discussion.

I concur!   The implications of the paper are actually rather disturbing; one of which being, that if we turn off the aerosols, we could get a prompt increase in temperature of as much as 1.8 degrees.

That will cause more havoc than I care to think about to deeply right now.

[crandles]

S Pansa,

Thanks for that paper. And thanks to everyone else for the discussion.

I concur!   The implications of the paper are actually rather disturbing; one of which being, that if we turn off the aerosols, we could get a prompt increase in temperature of as much as 1.8 degrees.

That will cause more havoc than I care to think about to deeply right now.

I also concur.

1.8C extra warming compared to 1.2C warming looks highly significant - 1.5*warming already occurred.

Rough eyeball of DMI graphs suggest to me about 5C warming for about a third of a year compared to 1958-2002 average which is 1.7C over a shorter period but presumably DMI's N of 80N weighted toward pole gets more arctic amplification than larger area shown on carbon brief map.

Probably pushing the interpretation rather a long way but .... if I take 5C warming for third of year and multipy by 1.5 to get a further 7.5C warming for the winter third of year. This would reduce the FDD by about 7.5C/25C or about 30%. 30% less ice volume at maximum reduces the volume from about 23000 km^3 to a little over 16000 km^3. We have melted 18000 km^3 in a season and this would rise if there was less ice to start. Not sure if we are going to fully turn off aerosols soon, but even a 75% turn off looks like it would be sufficient to get to an ice free day in September according to these rough calculations.

[jdallen]

<snippage>

Probably pushing the interpretation rather a long way but .... if I take 5C warming for third of year and multipy by 1.5 to get a further 7.5C warming for the winter third of year. This would reduce the FDD by about 7.5C/25C or about 30%. 30% less ice volume at maximum reduces the volume from about 23000 km^3 to a little over 16000 km^3. We have melted 18000 km^3 in a season and this would rise if there was less ice to start. Not sure if we are going to fully turn off aerosols soon, but even a 75% turn off looks like it would be sufficient to get to an ice free day in September according to these rough calculations.

It implies that with reduced Aerosols we could have an ice-free September with even a fairly 'Average' melt season and no additional CO2 or other climate forcing.

Rather bad news, that.

[ChrisReynolds]

crandles,

DMI uses reanalysis, the paper uses HADCRU. Reanalysis covers the arctic ocean,HADCRU does not. So this paper omits the warming over the ocean, which at the surface is due to loss of sea ice - less sea ice in summer, thinner ice in winter. So the higher DMI warming contains an element due to a feedback that may not be directly relevant.

[Michael Hauber]

The Najafi analysis uses the CMIP models for Arctic temperature change using three different forcing scenarios - mixed greenhouse gases, other anthropegenic forcings (i.e. anthro aerosols), and natural factors (volcanic aerosols and solar).  A scaling factor is then obtained for each run so that the three scenarios can be added together to get as close as possible to the HADCRU temperature record.  They find the best match is obtained when using scaling factors of almost exactly one for GHGs, and near 2 for other anthro.   In other words the models would match the historical temperature record much closer if the response to anthro aerosols was doubled. 

But how reliable is this result?  The author shows a chart with an error range on the scaling factor that is from about 1.2 to 2.5.  So assuming the error range is correctly calculated a scaling factor of 1 is not far outside the error range.  It is also interesting to note that the author notes that while aerosols in the Arctic have offset roughly 60% of the GHG warming, aerosols globally have offset  5% of GHG warming.  This result seems quite strange, and absent a mechanism to explain it I'm inclined to suspect that the method is just not accurate enough to tell the difference between 5% and 60%.  Having a look at the paper that Najafi et al quote I find that that they do use the same analysis method, and that the quoted range for aerosol offset is 0 to 0.5 degrees of warming out of 0.6 to 1.2 degrees.  That is on a global scale aerosols have offset between 0% and 83% of GHG warming.

[jai mitchell]

none of these temperature relationship analyses include the feedback effects of going below 1 Mkm^2 sea ice with significant sunshine still available before Sept. 21  .  With no regional ice left to melt in the CAB and large amounts of solar energy reaching the arctic oceans, regional air temperatures will rise to +15C.

[ChrisReynolds]

+15degC?

Perhaps. But by when? We all agree (I think) that the Arctic will become seasonally ice free. But that doesn't necessarily mean such intense warming. If we burn all the reserves of fossil fuels there seems a reaonable expectation of a perennial ice free state. But how long will that take?

[jai mitchell]

my calculations of regional increased solar insolation after the complete loss of sea ice in the CAB shows that these temperatures will rise rapidly to this level.  It is the remaining large volumes of ice in the region that are keeping temperatures near 0C during the melt season.

If we see a sept. minimum before 2020 then we will see these temperatures within 10 years or so, so by 2030.  Possibly earlier if aerosols continue to catch up.  Remember, there is a 10-year delay on maximum warming according to Ricke and Caldeira (2014).  This means a 10-year delay on the water vapor forcing feedback as well (for a single pulse of CO2 emissions).  Aerosols work to delay this water vapor feedback as well as it is temperature dependent.  This means there is significant regional forcing locked in to this region over the next several decades, even without additional sea ice loss.

[ChrisReynolds]

my calculations of regional increased solar insolation after the complete loss of sea ice in the CAB shows that these temperatures will rise rapidly to this level.  It is the remaining large volumes of ice in the region that are keeping temperatures near 0C during the melt season.

Can you detail your calculations? How do you calculate the temperature of the ocean under insolation? What mixing depth do you assume? What transfer of heat to the atmosphere do you assume (I'd love to know how to calculate atmospheric temperatures under such condtions)? Related to that last: How do you calculate winter warming given i) venting of heat from late season open waters, ii) heat flux through thinner ice? (Again - I don't know how to do that without using a column model - so a short cut would be great!)

[jai mitchell]

I am only talking about regional temperatures during the remaining days of high solar intensity, with some increased warming extending through October due to increased humidity.  I am not talking about annual averages.  I am not calculating energy loss during the winter months in a balance equation.  I already told you my position on the Tische paper's return to normal conditions after this kind of early sea ice loss event.

the calculation of the temperature response operating on the top 2 meters of ocean surface is clear, even more clear is the fact that a significant portion of this warming will be coming not just from incident solar radiation but also from off of the surrounding land surface. 

We will basically be seeing regional temperature increases in the CAB that are analogous to current Barrow Alaska temps in August.

[crandles]

Barrow has a good deal of land around it and land warms much more rapidly than ocean. Even then 46F looks like 14F above freezing not 15C, and that is before considering the daily low temp of 36F.

[jai mitchell]

I assume that you can find the specific heat capacities of air and sea water.

http://onlinelibrary.wiley.com/doi/10.1002/grl.50768/abstract

[jai mitchell]

Barrow has a good deal of land around it and land warms much more rapidly than ocean. Even then 46F looks like 14F above freezing not 15C, and that is before considering the daily low temp of 36F.

nor did I say that it was a daily average temperature, I said, "regional air temperatures will rise to 15C"  If you see the peak temperature of 60 degrees F then you get the peak air temperatures I was discussing.