Link of atmospheric warming "hiatus" to melting of arctic sea ice?

[dlen]

Hello Forists,

I did some calculations  the result of which is, that if the heat for the melting of the 8000 km³ ice in the past 18 years would have come from the atmosphere only, a mean temperature lag against expectations of -0.4 K would have occurred. This has to be put in relation to a real temperature lag of -0.1 to -0.2 K. So to me, it seems perfectly possible, that atmospheric heat transport has been changed in such a way as to accelerate arctic melting and cool the rest of the atmosphere.

The calculations can be found here:
http://remarksandobservations.wordpress.com/2013/11/15/global-warming-hiatus-it-might-be-the-ice/

What do You think of that?

[ggelsrinc]

Anyone who thinks sea ice melts according to temperatures has to think of things in superficial ways. It's a dynamic process and there is no hiatus, even if I hope for one. Right now that ASI is history in the making, because of what's been done, so enjoy your moment in the sun!   

[jdallen]

It no doubt contributed, Dien, but is still a small part of the story, energy wise.  As an energy sink, in raw terms, far more has been picked up by the world's oceans, which now have multiple pentajoules more sensible heat than just a few decades ago.  I have thought, and still think, that some of the supposed hiatus in measurements may be directly tied to heat exchange with water at depth.

The increase in trapped heat resulting from higher CO2 concentrations did not stop... Physics and chemistry lead me to that. It had to be picked up somewhere, and seawater is our largest heat buffer....

[ggelsrinc]

When I analyze something, I have a tendency to convert things to the simplest of terms. It's my understanding that "hiatus" means scientists believed global temperatures didn't increase as fast as they thought it would or as they expected.



Having been asked to explain the relationship of global warming to increases in CO2 has lead me to the obvious conclusion that an annual increase in CO2 is not going to correspond to an annual increase in global temperature. Global temperature is only a measurement of the skin of our planet. Decadal increases in CO2 show up in decadal increases in global temperatures, but annual increases don't. There are obviously other factors at play during shorter periods of time.

I've made a similar argument for global warming to your analysis many times, but made it with simple calculations in the form of this caveat. What happens to the energy to melt ice when the ice is all melted? I didn't just use ASI, but any ice for my example and kept it simple. Picture yourself standing on just one square foot of our planet over time and that square foot represents a micro-climate on Earth. If that square foot of micro-climate happens to be in an area where ice existed, like from a glacier, for example, what happens when that glacier retreats and the heat isn't going to melt ice that doesn't exist? The changes in that little micro-climate area influence and are influenced by all those little micro-climate areas around it and eventually the whole Earth.

Delta global T is only a representation of global warming, limited by it's construction. In the above image showing temperature trend, it's easy to focus on areas with increases, but try explaining those areas not showing a temperature trend increase! Places like the southeast US have a long history of being hot and humid and I wouldn't expect much change there. The data is complete enough to demonstrate something is happening against the trend in the Atlantic Ocean. That ocean has been cooling for millions of years and doesn't want to warm, or should I say will not warm quickly in short periods of time scales.

If heat isn't showing itself in global temperature changes, I'd say the oceans are the main reason. Climate patterns like ENSO and PDO are also major factors of variability or noise in the climate signal, but the oceans are the main driver, IMO. I've considered everything I know of that could possibly be a cause for the so-called "hiatus", but I think the order of magnitude of the oceans trumps them all. That doesn't mean your analysis isn't also important and everything else has it's importance in the summation of explanation, relative to their own scale.   

[dlen]

As an energy sink, in raw terms, far more has been picked up by the world's oceans, which now have multiple pentajoules more sensible heat than just a few decades ago.

You are right insofar as 0.9 of the global energy uptake goes into the oceans and only 0.1 into the atmosphere.

But this does not mean, that when the earth as a whole keeps on taking up energy unabatedly and the atmosphere takes up less of it, the resulting surplus energy has all to have gone into the oceans as well. A considerable part may have been transported into the arctic. How big this part is - I don't know.

What I do know is, that the atmospheric heat transfer into the arctic is surprisingly 5 times bigger than the heat transfer into the oceans (very roughly).

In Serreze - Arctic Energy Budget - J o Geophys. Res. Lett.  112 2007, p.12 , the mean power flow into the arctic is about 1,5 PW. From Levitus - World ocean heat content and thermosteric sea level change (0-2000), 1955-2010 - Geophys. Res. Lett. comes a heat flow into the oceans of only 0.3 PW.

So now how come the oceans take up 90 % of all the energy the whole earth takes up, when so much of it is taken to the arctic? 

The answer is simple: practically all of the energy taken to the arctic is radiated into space. Actually, top of atmosphere radiation even exceeds atmospheric heat transfer, because there is also the heat transfer by ocean currents.

The arctic is a gigantic power outlet radiating 1.7 PW into space.
To cause the additional perennial sea ice melting, only 0.004 PW was necessary, this is about 0.25 % of the heat transported to the arctic by the atmosphere!

Such a small possible increase in transport  power cannot be detected when putting up a regional energy balance, because it is a small difference of values polluted by much larger error margins.

Bottom line: it is not only possible, it seems to me very probable, that the "hiatus" missing energy went into ice melting.

[ggelsrinc]

Well, I don't think so.

I think the evidence points to the oceans and "hiatus" will not be a consideration once ENSO and PDO respond to their historical values. If the Earth warms too much, it's permanent El Nino times again, even with the Atlantic doing it's thing.

[RichardStamper]

There is a recent blog post at realclimate.org (http://www.realclimate.org/index.php/archives/2013/11/global-warming-since-1997-underestimated-by-half/) drawing attention to a paper that concludes that the "hiatus" doesn't really exist, but is an artefact of the relative lack of terrestrial temperature data inputs to the global models in certain regions of the globe, including the Arctic.  By using other sources of data, notably from satellites, to constrain the uncertainties in temperatures and interpolating intelligently (Kriging) they find that temperatures since the late-1990s have continued rising in line with the long-term trend. 

This isn't quite the same as saying that the "hiatus" is due to melting of Arctic sea ice - more like the melting of Arctic sea ice is a sign that the "hiatus" was never really there.

[jdallen]

Dien - as ggelsrinc was pointing out, the climate is not monolithic; "information"... That is to say heat...  Takes time do distribute through the system, and does so asymmetrically.  For example, the atmosphere provides a fraction of the energy uptake from insolation, but re radiates it at a much high rate than oceans.

The temperature is not so important as is the net increase in energy present in the system.  For that reason, the increase in enthalpy in the arctic is crucial.  At some point, that energy level will be sufficient such that the winter season heat loss will not be sufficient  enough to prevent spring and summer insolation from driving surface temperatures  above that required to retain a permanent ice pack. That threshold covers a very narrow range... About zero to minus two C.

When that happens, our climate will be in a transitional state heading towards an almost inevitable end point - a year round ice free arctic.

In geological terms, it will be a momentous event, as at this point in the suns life cycle, and absent of continental drift placing land mass at the pole, it is unlikely that the arctic ice cap will ever return.

[Shared Humanity]

It no doubt contributed, Dien, but is still a small part of the story, energy wise.  As an energy sink, in raw terms, far more has been picked up by the world's oceans, which now have multiple pentajoules more sensible heat than just a few decades ago.  I have thought, and still think, that some of the supposed hiatus in measurements may be directly tied to heat exchange with water at depth.

The increase in trapped heat resulting from higher CO2 concentrations did not stop... Physics and chemistry lead me to that. It had to be picked up somewhere, and seawater is our largest heat buffer....

I think you hit the nail on the head. The map of surface temperatures is simply evidence of this. Land surface temperatures are rising faster than the oceans because the mechanisms for dispersion of surface temperature rise are limited. The oceans mix and, thus. moderate the surface temperature increases. As mentioned on ASIB, this shows up as ocean expansion and sea level rise.

If we want to talk about "hiatus" then we should emphasize this is only with surface temperatures. The heating up is occurring, more rapidly than ever.

[ChrisReynolds]

Dlen,

Sorry but I disagree with your method.

Note that the specific heat content of air is specified at 1000 mbar and zero degC (STP), I'm not sure how this varies with temp/pressure, but much of the mass of the atmosphere is substantially away from STP.

You've taken an energy loss for the period 1995 to 2012(?) and applied it instananeously to temperature by calculating the effect of that net energy losses impact on temperature through specific heat capacity. Perhaps a better way of doing it would be to spreadsheet the energy calculation for each year 1995 to 2012, and apply each year's volume loss. It would be a better way for me to try my calcuations.

The way I've done it is to calculate the climate forcing implied by the energy loss (I've used the average loss from 1995 to 2012 for the monthly average volumes for April and September, average of those two losses). Physical constants for sea ice from Winton 2000 "A Reformulated Three-Layer Sea Ice Model"

Density of ice      905   kg/m^3
Latent Heat of Freezing      334000   J/kg
Loss of volume      7287   km^3
Loss of volume      7.287E+12   m^3

Having density latent heat of freezing and total volume I can work out the implied energy investment in Joules.

Implied energy investment      2.20264E+21   J
Timescale      18   Years
Timescale      567648000   seconds
Watts (j/s)      3.88029E+12   

Now having Watts, the power implied in melting the ice, I can work out the global forcing in Watts/m.

Area of Earth      510072000   km^2
Area of Earth      5.10072E+14   m^2
Forcing Watts/m      0.007607347   
Actually the forcing will be negative.
Forcing Watts/m      -0.007607347   

Now with the forcing I can work out the implied cooling with a climate sensitvity equivalent to 3degC for a doubling of CO2.

Hansen's Fast Climate Sensitivity      0.75   degC/W/m^2
Temperature rise      0.00570551   degC
 
Note that if you divide your calculated 0.4 degC by 18 years, you get 0.022degC, which  considering the very different methods of calculation, isn't too far from my 0.006, only a multiple of 4.

That the impact is so small doesn't surprise me, you just have to look at how small the year on year volume loss is, naively we can calculate this as 7287/18 = 404. With an annual range of volume of ballpark 15000 km^3 (actually a low end figure for recent years), the annual range is 37 times the annual loss of volume. Yet the sea ice only pegs temperature over the Arctic ocean at around zero, while temperatures over open ocean away from the ice reach +5degC in late season, and the temperatures over land soar in summer. With regional warming showing little more than a suggestion of levelling since (IIRC) 2005.

Furthermore as you've read Kosaka and Xie you'll appreciate that they find a winter cooling, with a patch of warming that isn't in reality warming but is cooling - lending support to Cohen's work on cold boreal winters.

[seaice.de]

Fig. 8. Contributions to planetary energy imbalance by processes other than ocean heat uptake.

http://www.atmos-chem-phys.net/11/13421/2011/acp-11-13421-2011.pdf

[dlen]

There is a recent blog post at realclimate.org (http://www.realclimate.org/index.php/archives/2013/11/global-warming-since-1997-underestimated-by-half/) drawing attention to a paper that concludes that the "hiatus" doesn't really exist, but is an artefact of the relative lack of terrestrial temperature data inputs to the global models in certain regions of the globe, including the Arctic. 
...
This isn't quite the same as saying that the "hiatus" is due to melting of Arctic sea ice - more like the melting of Arctic sea ice is a sign that the "hiatus" was never really there.

Thinking more and more about it, a picture forms, in which the accelerated ice loss is kind of a collateral damage of atmospheric heat (AH) beeing transported more efficiently into the arctic.

This basically coincides with the paper You kindly mentioned. If the "hiatus" vanishes when we take the arctic properly into account, the term can still be used in the restricted meaning as "hiatus of the non-arctic atmospheric temperature".

The reason for seeing things differently is, that the heat flow from AH transfer does mostly not go into the ice sheet, but goes past and is radiated directly into space. The heat flow through the ice sheet by contrast comes from the water beneath it.

So, when, because of more efficient AH transfer, more AH reaches the arctic, it's the sea -> ice sheet -> air heat-flow that starts happening on a higher temperature level, which in turn erodes the ice.

OTOH, with the melting ice eating up some of the oceanic heat coming from below, it helps keeping temperature lower, stabilizing  the temperature difference between the arctic and the rest of the world and thus supports the AH transfer efficiency (This latter conclusion relies on the hypothesis, that transport efficiency always rises with rising temperature difference, which is not proven.)
 
This AH transfer enhancement effect of the melting (when it exists)  can at least with some right be called one "cause" of the "hiatus" in the latter sense.

[ggelsrinc]

ASI also traps heat. Couldn't we cool our planet in our atmosphere by making ASI and wouldn't that warm that cold ocean?

[Bob Wallace]

This may bring a bit of discomfort to those who have been taking solace in the "pause".

The planet may be warming much more – and much faster – in recent years than many experts have believed, according to a new study released this week by the U.K.-based Royal Meteorological Society.

Prepared by British and Canadian researchers, the study reports that the rise in global temperatures over the past 15 years has been significantly underestimated due to gaps in temperature data around the world, largely in Earth's polar regions.

“It turns out that we only have surface measurements over about 84 percent of the globe," said Weather Underground's Dr. Jeff Masters, noting that there are no direct measurements of temperatures in places like the Arctic even today, especially across its vast stretches of sea ice.

Relying on weather station data to determine globally averaged temperature trends, as RealClimate.org points out, has a long-acknowledged weak spot: temperature records in places like Antarctica began only in the 1950s, and they're extremely limited even today in much of the Arctic and parts of Africa.

Leaving out these parts of the planet means that the most widely-used global temperature data sets – like the U.K. Met Office's Hadley Center observations and NOAA's National Climatic Data Center – estimated that these regions were warming at the same rate as the rest of the world.

Using satellite data to fill in those gaps, the study found that global surface temperatures have been warming 2 1/2 times faster over the past 16 years than previously believed, calling into dispute the widely-reported global warming "pause," the observed slowdown in the rise of global temperatures since 1997 despite steadily-increasing amounts of greenhouse gases in the atmosphere.

“When you do this technique, it turns out that this so-called slowdown or hiatus in the global surface temperature, it disappears," said Masters. "We don’t see any slowdown at all in the rate of global warming with this new estimate of the missing data in the polar regions.

http://www.wunderground.com/news/no-global-warming-pause-planet-warming-much-faster-thought-20131114

[ggelsrinc]

This may bring a bit of discomfort to those who have been taking solace in the "pause".

The planet may be warming much more – and much faster – in recent years than many experts have believed, according to a new study released this week by the U.K.-based Royal Meteorological Society.

Prepared by British and Canadian researchers, the study reports that the rise in global temperatures over the past 15 years has been significantly underestimated due to gaps in temperature data around the world, largely in Earth's polar regions.

“It turns out that we only have surface measurements over about 84 percent of the globe," said Weather Underground's Dr. Jeff Masters, noting that there are no direct measurements of temperatures in places like the Arctic even today, especially across its vast stretches of sea ice.

Relying on weather station data to determine globally averaged temperature trends, as RealClimate.org points out, has a long-acknowledged weak spot: temperature records in places like Antarctica began only in the 1950s, and they're extremely limited even today in much of the Arctic and parts of Africa.

Leaving out these parts of the planet means that the most widely-used global temperature data sets – like the U.K. Met Office's Hadley Center observations and NOAA's National Climatic Data Center – estimated that these regions were warming at the same rate as the rest of the world.

Using satellite data to fill in those gaps, the study found that global surface temperatures have been warming 2 1/2 times faster over the past 16 years than previously believed, calling into dispute the widely-reported global warming "pause," the observed slowdown in the rise of global temperatures since 1997 despite steadily-increasing amounts of greenhouse gases in the atmosphere.

“When you do this technique, it turns out that this so-called slowdown or hiatus in the global surface temperature, it disappears," said Masters. "We don’t see any slowdown at all in the rate of global warming with this new estimate of the missing data in the polar regions.

http://www.wunderground.com/news/no-global-warming-pause-planet-warming-much-faster-thought-20131114

I don't think we have 84% coverage of the Earth's surface, even with satellite temperature measurements, which need to be calibrated by instrumental measurements. The last time I checked, the reality was we are getting better measuring temperatures in areas of our world, but determining delta T for an area always requires accurate past measurements during a base period. I remember the Russians complaining, because large areas weren't included in the Met data, so Denialistas seized on that point following the email hacking fiasco, claiming the data was cherry picked, as if a cherry still existed on the tree after they were finished.

We only know what we know.

When climate patterns in the Pacific start favoring warming our Earth, the word hiatus will drop out of vogue. 

[dlen]

Sorry but I disagree with your method.

You've taken an energy loss for the period 1995 to 2012(?) and applied it instananeously to temperature by calculating the effect of that net energy losses impact on temperature through specific heat capacity. ...

The way I've done it is to calculate the climate forcing implied by the energy loss .

...

Note that if you divide your calculated 0.4 degC by 18 years, you get 0.022degC, which  considering the very different methods of calculation, isn't too far from my 0.006, only a multiple of 4.

...

I've no problem with that. My calculation is really rough and coarse and only showed, that the energy consumed by the ASI melting is really a factor to recon with. And a bad hypothesis, that triggers an insightful process is often how research and understanding works.

If we map the energy flow system by a network of capacities and resistances in steady state, and change somewhere something, there will be a transient state, after which there will be reached a new steady state. For the new steady state, only the resistances are of importance. This is Your approximation using the climate forcing and its effect on surface temps, which reflect the complex reaction by just one number: 0.75 K/(W/m²).

My approximation was more or less ignoring all resistances and steady state flows and taking only the capacities into account.Although the forcing approach is better then the pure heat-content-approach, it is not perfect. What You did is to add an artificial heat flow in the flow network.



The line thickness is roughly proportional to the thermal power flow they represent.

My point is, that it makes a difference for the temperature response, where You put in the extra heat drain.

Approximate the network further by two thermal resistors:

R1 from input point to the arctic (via atmosphere and ocean)
R2 from arctic to output point ( space)  via vertical convection and radiation,
a thermal power input P.

Then the total temperature against space is given by  P (R1 + R2).
So if I add some minor power p, the temperature difference is given by p (R1 + R2).
In our case, p is negative.
If I add p in the arctic, the temperature difference is only p R2.
This means, that for heat flows added somewhere inside the network, the sensitivity is lower.

Also, there is the question of the time constant. You wrote "hansens fast climate sensitivy". Would You be so kind to give me a link? What does "fast" mean?

So both envelope calculations are somewhat flawed. Interestingly, Hansen et al.: Earth's Energy Imbalance and Implications.. (seaice.de's link from above) delivers about the double value of Your forcing stemming from ASI melt. They took into accout not only the melting enthalpy, but also a temperature rise from -10 to +15 °C. But this doesn't make up for the difference.

In the picture drawed by the paper, the heat uptake by ASI melt is only roughly 1/4 of the total non-ocean heat uptake, with a distinct rise in the "hiatus" period, though. The other components are either wildly oscillating or show a slower, more continuous rise, while ocean heat uptake seems to have decreased in the period in question! So IMO the importance of ASI melt for the rest-of-world-hiatus is supported.

[ChrisReynolds]

Dlen,

But you've still calculated the total impact from 1995 to 2012 as an instantaneous impact, I don't think this is right. My calculation is the average from 1995 to 2012, taking into account the variation in figure 8d, and the fact that I'm using PIOMAS not the Rothrock data in my opinion puts my calculation more in line with Hansen than even yours divded by 18 (years). But the 0.4degC is overstated by at least a multiple of 18 in terms of your method of calculation.

Hansen fast climate sensitivity is described in this paper.
http://arxiv.org/ftp/arxiv/papers/1211/1211.4846.pdf
Which was later formally published here:
http://rsta.royalsocietypublishing.org/content/371/2001/20120294

[dlen]

Well Chris,

as I wrote my calculation completely ignored any steady state flows and is outrageously primitive. Its only objective was to show, that the energy necessary to melt the ice is bigger than the the energy difference in the atmosphere between model and measurement.

In a system, where all other energy flows remain constant, it makes no difference wether I transfer this energy all at once or in a steady stream over a certain period of time.

Of course the earth is not such a system. When I drain energy at one point, all other flows adjust a little bit until a new balance is reached.

So the result of my calculation can be considered as an upper bound for the temperature response to energy drain from melting. With a factor of 80 or so above the steady state value it is admittedly a very generous one. Nevertheless, if the outcome would have been, that the energy of melting was much smaller than the energy lag in the atmosphere, this mechanism would have been ruled out from the beginning.