RealClimate

Climate science from climate scientists...

You are here: Home / Climate Science / Climate modelling / The underestimated danger of a breakdown of the Gulf Stream System

The underestimated danger of a breakdown of the Gulf Stream System

by

A new model simulation of the Gulf Stream System shows a breakdown of the gigantic overturning circulating in the Atlantic after a CO2 doubling.

A new study in Science Advances by Wei Liu and colleagues at the Scripps Institution of Oceanography in San Diego and the University of Wisconsin-Madison has important implications for the future stability of the overturning circulation in the Atlantic Ocean. They applied a correction to the freshwater fluxes in the Atlantic, in order to better reproduce the salt concentration of ocean waters there. This correction changes the overall salt budget for the Atlantic, also changing the stability of the model’s ocean circulation in future climate change. The Atlantic ocean circulation is relatively stable in the uncorrected model, only declining by about 20% in response to a CO2 doubling, but in the corrected model version it breaks down completely in the centuries following a CO2 doubling, with dramatic consequences for the climate of the Northern Hemisphere.

The potential instability of the Atlantic Meridional Overturning Circulation or AMOC – commonly known as the Gulf Stream System – has been a subject of research since the 1980s, when Wallace Broecker warned in an essay in Nature of Unpleasant Surprises in the Greenhouse. The reason for this was growing evidence of abrupt climate changes in the history of the Earth due to instability of Atlantic currents.

Fig. 1 Schematic of the Atlantic ocean circulation (simplified). In red the relatively warm surface flow is seen, in blue the cold deep water flow. The northward surface flow and southward deep flow together make up the Atlantic Meridional Overturning Circulation (AMOC), popularly dubbed Gulf Stream System. Image by S. Rahmstorf (Nature 1997), Creative Commons BY-SA 4.0. 

Why the AMOC has a tipping point

The basic physical mechanism of this instability was already described by Henry Stommel in 1961. The freshwater balance (precipitation minus evaporation), which determines the salt content, is central to this. Freshwater continually flows into the northern Atlantic through precipitation, rivers and ice-melting. But supply of salty waters from the south, through the Gulf Stream System, balances this. If however the current slows, there is less salt supply, and the surface ocean gets less salty. This fresher water is lighter than saltier water and therefore cannot sink into the depths so easily. Since this sinking – the so-called deep water formation – drives the Gulf Stream System, the current continues to weaken. There is a critical point when this becomes an unstoppable vicious circle. This is one of the classic tipping points in the climate system.

However, it’s still unclear where exactly this tipping point is. Most models show a significant slowdown in the Gulf Stream System by 20% to 50% in typical global warming scenarios up to the year 2100, but do not exceed the tipping point that would lead to its collapse. However, there is a large spread between different models – which is not surprising since the stability of the Atlantic flow depends on a subtle balance in the salinity and thus also in the freshwater budget, which is only inaccurately known. In addition, there have long been serious indications that the models are not only inaccurate, but perhaps all systematically biased towards an exceedingly stable AMOC. We discussed these papers in a review article in PNAS in 2009.

What makes the new study different?

According to lead author Wei Liu, the starting point of the new study was my paper from 1996 on the relationship between the freshwater balance and stability of the flow. Back then I showed how to determine the stability of the AMOC from an analysis of the freshwater transport in the Atlantic at 30° south. The decisive factor is whether the AMOC brings freshwater into the Atlantic basin or whether it exports it (in the latter case, working to increase salinity in the Atlantic). My article ended with the suggestion to clarify this from observational data. That was later done by colleagues from Holland (Weijer et al. 1999). Several studies followed which performed this diagnosis for different climate models (e.g., Pardaens et al. 2003, de Vries and Weber 2005, Dijkstra 2007, Drijfhout et al. 2010, Hawkins et al. 2011). According to the observational data, the AMOC is exporting freshwater, which is why freshwater will accumulate in the Atlantic when the AMOC breaks down. That is precisely the instability described by Stommel 1961 and Broecker 1987. In the models, on the other hand, the AMOC in most cases imports freshwater, so the flow is fundamentally stable there. The differences in AMOC stability between different models cannot be understood without the fundamental criterion of whether the AMOC imports or exports freshwater, and by what amount. Liu et al. 2014 have identified a known common bias in all coupled climate GCMs without flux adjustments, the “tropical bias”, which makes them import freshwater in contrast to what observations show for the real ocean. A model bias towards stability is also consistent with the fact that most models underestimate the cooling trend observed in the subpolar Atlantic, which is indicative of an ongoing significant AMOC weakening, as we have argued (Rahmstorf et al. 2015).

The new study attempts to correct this model deficit by modifying the freshwater exchange at the sea surface in a model by using a so-called flux correction (which also involves the heat exchange, but this should be secondary). As a result, the salinity distribution in the ocean of the model for today’s climate is in better agreement with that of the real ocean. This is an important criterion: while precipitation and evaporation over the oceans are difficult to measure and therefore only very imprecisely known, we have detailed and precise information about the salinity distribution from ship measurements. Apart from the improved salinity distribution, this correction has no significant influence on the model climate for the present.

And now the result …

With both model variants – with and without the subtle correction of the salinity distribution – an experiment was performed in which the amount of CO2 in the air was doubled. The reaction of the Atlantic circulation is shown in the following graph. Without correction, the AMOC once again proves to be very stable against the massive disturbance. With the correction, in contrast, the flow breaks down in the course of about 300 years. It has lost a third of its strength after 100 years.


Fig. 2 Time series of the Atlantic flow (AMOC) in the two model variants: without correction (blue) and with correction (orange). In model year 201, the CO2 concentration in the model is doubled and then left at this level. Source: Liu et al., Science Advances 2016.

As expected, the breakdown of the heat-bringing Gulf Stream System leads to a cooling in the northern Atlantic, as shown in Figure 3. Land areas are also affected: besides Greenland and Iceland mainly Great Britain and Scandinavia.


Fig. 3 Temperature change in the winter months (DJF), 300 years after CO2 doubling in the experiment. Due to the almost completely extinct Atlantic flow, the northern Atlantic region has cooled significantly. Source: Wei Liu, with permission.

This new study is certainly not the last word on this important question. Compared to the measured data the correction appears to be somewhat too strong – the adjusted model version might therefore be too unstable. As computing time is scarce and expensive, the CO2 concentration in the experiments was abruptly doubled, rather than gradually ramped up in a more realistic emission scenario. The experiment was carried out with only one climate model; for robust conclusions, one usually waits until a series of models shows consistent results. (However, consistent with the new results two earlier climate GCMs and a number of simpler models have shown an AMOC that exports freshwater and is bistable, i.e. could potentially pass a tipping point and break down, as discussed by Liu et al. 2014.)

Also, no meltwater influence from the dwindling continental ice on Greenland was taken into account, which could additionally weaken the flow. On this topic, only three weeks ago a new study was published (Bakker et al. 2016) comparing future warming scenarios, once with and once without consideration of the influx of Greenland meltwater. (An emulator was used for this study; that is a highly simplified computer model that reproduces the results of complex circulation models in a time-saving way, so that many experiments can be performed with it.) With unmitigated emissions (RCP8.5 scenario), the Gulf Stream System weakens on average by 37% by the year 2300 without Greenland melt. With Greenland meltwater this doubles to 74%. And a few months ago, a study with a high-resolution ocean model appeared, suggesting that the meltwater from Greenland is likely to weaken the AMOC considerably within a few decades (Böning et al. 2016 – as we reported).

There are, therefore, two reasons why thus far we could have underestimated the risk of a breakdown of the Gulf Stream System. First, climate models probably have a systematic bias towards stable flow. Secondly, most of them do not take into account the melting ice of Greenland. As the new studies show, each of these factors alone can lead to a much stronger weakening of the Gulf Stream system. We now need to study how these two factors work together. I hope these worrying new results will encourage as many other research groups as possible to pursue this question with their own models!

Weblinks

Washington Post: Scientists say the global ocean circulation may be more vulnerable to shutdown than we thought

Climate Central: Potential for Collapse of Key Atlantic Current Rises

The Verge: Climate change may shut down a current that keeps the North Atlantic warm

The Atlantic: The Atlantic Ocean and an Actual Debate in Climate Science

Video lecture on the Gulf Stream System

More on the Gulf Stream System slowdown at RealClimate

Q & A about the Gulf Stream System slowdown and the Atlantic ‘cold blob’

AMOC slowdown: connecting the dots

What’s going on in the North Atlantic?

A hypothesis about the cold winter in eastern North America

Blizzard Jonas and the slowdown of the Gulf Stream System

References

  1. W.S. Broecker, "Unpleasant surprises in the greenhouse?", Nature, vol. 328, pp. 123-126, 1987. http://dx.doi.org/10.1038/328123a0
  2. H. STOMMEL, "Thermohaline Convection with Two Stable Regimes of Flow", Tellus, vol. 13, pp. 224-230, 1961. http://dx.doi.org/10.1111/j.2153-3490.1961.tb00079.x
  3. T.M. Lenton, H. Held, E. Kriegler, J.W. Hall, W. Lucht, S. Rahmstorf, and H.J. Schellnhuber, "Tipping elements in the Earth's climate system", Proceedings of the National Academy of Sciences, vol. 105, pp. 1786-1793, 2008. http://dx.doi.org/10.1073/pnas.0705414105
  4. M. Hofmann, and S. Rahmstorf, "On the stability of the Atlantic meridional overturning circulation", Proceedings of the National Academy of Sciences, vol. 106, pp. 20584-20589, 2009. http://dx.doi.org/10.1073/pnas.0909146106
  5. S. Rahmstorf, "On the freshwater forcing and transport of the Atlantic thermohaline circulation", Climate Dynamics, vol. 12, pp. 799-811, 1996. http://dx.doi.org/10.1007/s003820050144
  6. W. Weijer, W.P.M. de Ruijter, H.A. Dijkstra, and P.J. van Leeuwen, "Impact of Interbasin Exchange on the Atlantic Overturning Circulation", Journal of Physical Oceanography, vol. 29, pp. 2266-2284, 1999. http://dx.doi.org/10.1175/1520-0485(1999)029<2266:IOIEOT>2.0.CO;2
  7. A.K. Pardaens, H.T. Banks, J.M. Gregory, and P.R. Rowntree, "Freshwater transports in HadCM3", Climate Dynamics, vol. 21, pp. 177-195, 2003. http://dx.doi.org/10.1007/s00382-003-0324-6
  8. P. de Vries, and S.L. Weber, "The Atlantic freshwater budget as a diagnostic for the existence of a stable shut down of the meridional overturning circulation", Geophysical Research Letters, vol. 32, 2005. http://dx.doi.org/10.1029/2004GL021450
  9. H.A. Dijkstra, "Characterization of the multiple equilibria regime in a global ocean model", Tellus A, 2007. http://dx.doi.org/10.3402/tellusa.v59i5.15173
  10. S.S. Drijfhout, S.L. Weber, and E. van der Swaluw, "The stability of the MOC as diagnosed from model projections for pre-industrial, present and future climates", Climate Dynamics, vol. 37, pp. 1575-1586, 2010. http://dx.doi.org/10.1007/s00382-010-0930-z
  11. E. Hawkins, R.S. Smith, L.C. Allison, J.M. Gregory, T.J. Woollings, H. Pohlmann, and B. de Cuevas, "Bistability of the Atlantic overturning circulation in a global climate model and links to ocean freshwater transport", Geophysical Research Letters, vol. 38, pp. n/a-n/a, 2011. http://dx.doi.org/10.1029/2011GL047208
  12. W. Liu, Z. Liu, and E.C. Brady, "Why is the AMOC Monostable in Coupled General Circulation Models?", Journal of Climate, vol. 27, pp. 2427-2443, 2014. http://dx.doi.org/10.1175/JCLI-D-13-00264.1
  13. S. Rahmstorf, J.E. Box, G. Feulner, M.E. Mann, A. Robinson, S. Rutherford, and E.J. Schaffernicht, "Exceptional twentieth-century slowdown in Atlantic Ocean overturning circulation", Nature Climate Change, vol. 5, pp. 475-480, 2015. http://dx.doi.org/10.1038/NCLIMATE2554
  14. P. Bakker, A. Schmittner, J.T.M. Lenaerts, A. Abe‐Ouchi, D. Bi, M.R. van den Broeke, W. Chan, A. Hu, R.L. Beadling, S.J. Marsland, S.H. Mernild, O.A. Saenko, D. Swingedouw, A. Sullivan, and J. Yin, "Fate of the Atlantic Meridional Overturning Circulation: Strong decline under continued warming and Greenland melting", Geophysical Research Letters, vol. 43, 2016. http://dx.doi.org/10.1002/2016GL070457
  15. C.W. Böning, E. Behrens, A. Biastoch, K. Getzlaff, and J.L. Bamber, "Emerging impact of Greenland meltwater on deepwater formation in the North Atlantic Ocean", Nature Geoscience, vol. 9, pp. 523-527, 2016. http://dx.doi.org/10.1038/ngeo2740

Reader Interactions

70 Responses to "The underestimated danger of a breakdown of the Gulf Stream System"

  2. This is of significant practical concern to me, as my brother’s house in Portsmouth VA USA is roughly 1.5 meters above mean high tide.

    Here’s what I think I’ve heard. A brief 2012 35% slowdown in the Gulf Stream resulted in an immediate 4 inch rise in US East Coast sea level. And, a complete and permanent cessation of the AMOC would result in between a meter and a meter and a half of sea level rise along the US East Coast.

    That, on top of projected sea level trends, would make low-lying areas of the Virginia Tidewater region uninhabitable, I think.

    My question is whether the Gulf Stream will likely exhibit “start-and-stop” behavior as it slows. (As in the 2012 event.) And, whether the models that are being used can offer any information about that or not.

    Let me make my point crudely. If you project a 50% decline in Gulf Stream flow, I think it matters to me whether that’s a) a 50% decline in each of 12 months, or b) zero flow for six months, followed by a re-start at 100% flow for six months. I think flooding damage is likely to be much higher under b).

    And then, markets being what they are, if we see even one such significant and prolonged “stutter” in the Gulf Stream, it will probably become difficult to sell a US East Coast house that sits at 1.5 meters above mean high tide.

    So my general question is whether the Gulf Stream flow is expected to become more erratic as it slows in response to global warming.

  3. They may be altering the “reaction” of the model by applying flux adjustments, and given that this is only one model, I do not find their results to be that compelling. Several models in CMIP3 used flux adjustments (and so presumably had reduced state bias) and a reasonable number of CMIP5 models indicate that they are currently bistable (likely capable of a maintaining a collapsed state). However, only very few other models show that warming could cause a collapse of the AMOC, and only under fairly extreme warming (like RCP8.5). It does not mean that a rapid collapse is not possible – especially if Greenland melt is included – just unlikely given what we can glean from models so far.

    [Response: Hi Michael, good that you stop by! Have you had a look at Fig. 1 of Wei Liu’s paper? It shows that compared to observational estimates, all CMIP5 models have an AMOC that is a lot too stable, judging from the AMOC stability indicator. See also his previous 2014 paper. So I don’t think it is correct to just treat all models as equally likely and say that if only one model shows a collapse than this is unlikely to be correct. Rather, I think there is good evidence that all CMIP5 models are biased towards a too stable AMOC so I’d say it is likely that these are all wrong on this particular issue. -stefan]

  6. With CO2 doubling and Methane being added by leaps and bounds and Nitrous Oxide just starting a massive ramp-up, I doubt that these projections can be valid beyond the short term. I see no mentioning of the fact that way over 90 percent of the Arctic Sea Ice will not be there to melt and produce large amounts of freshwater content. The Greenland melt should continue, but as the ice melts in coverage area (more land is exposed) the melting ice run off should decrease. Nobody mentioned the stability of the salt content of the oceans, but with all that water vapor going into the sky I would expect salt content to go up. Is that written into the projections for 2100?

    [Response: What goes up must come down… What evaporates will rain down within a few days (I think ten on average). Moisture content of the atmosphere goes up a bit with global warming – true – but that is a negligible amount of water loss from the ocean’s point of view (worth just a few millimeters of sea level). A bigger effect on global ocean salinity is the continental ice loss, which dilutes the ocean water. You can work this out easily: average depth of the global ocean is ~3700 meters. So if you melt enough ice to add 120 meters to sea level (which happened at the end of the last Ice Age) you decrease salinity by approximately 120/3700 = 3 %. That is about 1 salinity unit (out of a mean ocean water salinity of 35 psu). That is, during the Last Glacial Maximum the global ocean was about 1 psu saltier than now. If we add ten more meters to sea level by melting ice in the coming centuries, that would reduce mean ocean salinity by about 0.1 psu. This reduction in the mean is not an issue – this freshwater pooling in the surface ocean in certain regions is the problem. -stefan]

  7. The Gulf Stream keeps North East USA and North West Europe warmer than they should be.
    If this heat belt slows the resultant lowering in temperature will be disruptive to both societies.
    I do not think it is a “IF” I feel it is going to happen with the amount of fresh water coming off the ice sheets and melting sea ice.
    Perhaps even the diluted sea water may be cooler and more dense than further south but this inherently must slow the drag effect that pulls up the warmer water towards this area.
    Perhaps i did not say that very well i mean that the sink of dense saline cold water will be less when the salinity is lessened.
    The outcome would be a colder situation for both sides of the North Atlantic.
    This of course will feed nicely into the people who regard the world as their back yard.
    Trying to explain this to people who are eyes closed is not very easy i feel

  8. Russell, ‘oogled your question and this popped up:

    https://www2.ucar.edu/atmosnews/news/18283/ncar-develops-method-predict-sea-ice-changes-years-advance

    Decadal prediction relies on the idea that some natural variations in the climate system, such as changes in the strength of ocean currents, unfold predictably over several years. At times, their impacts can overwhelm the general warming trend caused by greenhouse gases released into the atmosphere by humans.

    Yeager’s past work in this area has focused on decadal prediction of sea surface temperatures. A number of recent studies linking changes in the North Atlantic ocean circulation to sea ice extent led Yeager to think that it would also be possible to make decadal predictions for Arctic winter sea ice cover using the NCAR-based Community Earth System Model…

  10. I would think the modeled cooling in southern Greenland would greatly reduce the melting of the ice sheet. In that sense at least it ought to be a stabilizing influence.

    [Response: No Greenland melting whatsoever is included in these simulations. That is a far more drastic “stabilising influence” than the fact that once the AMOC has largely collapsed, melting from Greenland might be reduced. Once the AMOC is past its tipping point, reduced melting does not help anyway. -stefan]

  11. More:

    https://www.sciencedaily.com/releases/2012/06/120625162907.htm

    study published June 25 by the Proceedings of the National Academy of Sciences, the Greenland ice core drifts notably from other records of Northern Hemisphere temperatures during the Younger Dryas, a period beginning nearly 13,000 years ago of cooling so abrupt it’s believed to be unmatched since.

    … they just couldn’t find a lever in the model that would simulate a Younger Dryas that matched the Greenland ice cores.

    “You can totally turn off ocean circulation, have Arctic sea ice advance all the way across the North Atlantic, and you still will have a warmer climate during the Younger Dryas than the Oldest Dryas because of the carbon dioxide,” Carlson says.

  12. that’s cited to:

    Zhengyu Liu, Anders E. Carlson, Feng He, Esther C. Brady, Bette L. Otto-Bliesner, Bruce P. Briegleb, Mark Wehrenberg, Peter U. Clark, Shu Wu, Jun Cheng, Jiaxu Zhang, David Noone, and Jiang Zhu. Younger Dryas cooling and the Greenland climate response to CO2. Proceedings of the National Academy of Sciences, June 25, 2012 DOI: 10.1073/pnas.1202183109

  13. They always talk about cooling in NE US & N Europe due to the slow down or breakdown, but I’m wondering what would happen to the Gulf of Mexico? Would that get hotter?

  14. Do you guys think any of us will be around in 300 years? It’s doubtful.

    [Response: I sure doubt I will be… -stefan]

  15. Two thoughts:

    1: Any property at/near sea level is a risky investment – values will decline long before most properties are under water;

    2: Uncertainty (in this case, over AMOC) is not our friend – the outcome could be worse than thought, rather than better. All the more incentive to minimise risk by returning to Holocene-like climate conditions ASAP. Not that we will…

  16. How do short term climate forcing agents (like methane and black carbon) contribute to the danger of triggering a tipping point in the Gulf Stream System?

    We are sometimes told that the temperature increase from the short-lived agents dissipates relatively quickly so their effects are only important when we are close to the “peak temperature”.

    I would be interested to know where the extra heat goes that is trapped by the agents: radiated into space, ice melt, warmer oceans or somewhere else?

    Do any of these increase the danger of triggering a Gulf Stream System tipping point? If so, how much?

  17. I do wonder if the role of the Gulf Stream is overstated when it comes to the UK’s temperate climate. I have heard it stated in the past that if the Gulf Stream shut down, it would lead to another ice age. My thought is that the UK’s temperate climate is primarily due to SWly prevailing winds blowing from the relatively warm ocean, as opposed to places on the eastern side of a large continent which will frequently receive cold polar continental airmasses in winter, because the mid-latitude westerlies will be blowing from a cold continental interior. The Gulf Stream is a second order effect in my opinion. Compare the mean daily max and min temperatures over each month of Seattle and London. Seattle has cooler sea surface temperatures in the Pacific Ocean upstream of the W/SW prevailing winds.

    [Response: The model includes all those effects of course, so Fig. 3 in a way answers your question about how strong the AMOC effect is. Compare London there with places on the same latitude at the Pacific. I see about 3 °C difference. It’s bigger for Ireland and Scotland, and much bigger for Iceland. -stefan]

  18. This is an interesting, disturbing finding.

    But I have a fundamental question: Why is the AMOC the same as the “Gulf Stream System”? Isn’t there also a significant component of the latter that is wind driven? Is this also breaking down in this simulation?

    [Response: I am using terminology here from the MPI in Hamburg (as linked with the term under Fig. 1). It is a question of translating the more specific and complex scientific terminology to something a general audience can relate to; they write:

    The somewhat complicated terminology of the various parts can be avoided by use of the general term ‘Gulf Stream System’.

    And of course the Gulf Stream itself does not break down as it is largely wind driven – to show that component, my schematic diagram also includes the wind-driven subtropical gyre, to which the Gulf Stream is the western boundary current. -stefan

  19. Barbara asked: “Do you guys think any of us will be around in 300 years? It’s doubtful…”

    That’s a disturbing question, hinging on what’s meant by “us.” Of course it’s unlikely to umpty-ump decimal places that any of the individuals reading this thread will be. So are you suggesting that H. Sap–an abbreviation that seems more apt than ever, somehow–is at risk of extinction? Inquiring minds and all that…

  20. @17: It is important to understand that the NAC is largely wind-driven, in other words, the two go together, which means that less Gulf Stream (roughly) = less SW wind reaching NW Europe.

    The comment about ‘another ice age’ sounds too much like the Daily Mail, not a reliable source for anything about climate. In any case, it’s almost certainly media hyperbole rather than a scientific finding.

    One thing the graphic fails to reference is the baseline of the temperature changes – is that relative to present temperatures, or relative to the temperatures as they are likely to be down the line? It makes a big difference.

    As with all analysis of the AMOC, the critical unanswerable is the timeline, and the way in which the patterns change. If the process is progressive, then impacts like the Winters of 2010-2011 in the UK will repeat with increasing frequency until the background GW overtakes it. UKCIP 2009 had this down as between the 2080’s and 2100. It doesn’t feel that a lot has changed, here, apart from the degree of confidence that a slowdown/shutdown is inevitable.

  21. At this moment, we have open water at the Bearing Straight. PIOMAS ice volume is being analyzed for the December average but it appears to be coming in around -8.5% below the previous record low.

    As the NH Summer insolation returns to the upper latitudes, this lack of ice will greatly amplify regional albedo impacts. This effect is going to rapidly increase as China continues to reduce its SO2 emissions and sea ice volume drives further ice declines. I estimate the first September ice free condition sometime around 2021 +/- 2 years and no later than 2025.

    Due to the effect of returning solar insolation and retreating ice edge boundaries, this increased solar absorption is very great. A 2013 Caldeira paper showed ~3.0 W/m^2 globally averaged forcing for a year-around ice free state. I estimate that an ice free condition of June 21 will be equal to the total forcing produced by a doubling of CO2 WITH LR/WV feedbacks. (~2.3 W/m^2).

    However, this effect will NOT be globally averaged but will instead be directed wholly at the Arctic Ocean. I expect this to occur sometime around 2065 but will very likely happen much more rapidly if recent studies of non-linearity of ECS are correct and ECS is closer to 4.5 vs. current 3.0.

    In this new regime, with a complete absence of sea ice and snow in the Northern Hemisphere, with rapid warming of the arctic region due to increased solar absorption, a jump in regional temps will occur. This rapid warming impact will greatly increase the meridional transport of 1000Hpa tropical water vapor into the region, leading to rainfall on the slopes of Greenland up to 600 meters elevation.

    At this point we will begin to see the Greenland pulse, on a scale of the Agassiz event. Though the dynamics described above may be sufficient to halt AMOC flow before that.

    p.s. the current slowing of AMOC appears to be a factor in the current jumps in both CO2 and CH4 atm. abundances. I estimate AMOC is responsible for ~5-8% of the total oceanic sink.

  22. Thank you. The video is amazingly accessible to this maths-challenged observer who loves visuals.

    I’m in New England, and note that we will be “warmer” which contradicted my uneducated impression. While weather is not climate, it is peculiar to be so warm in winter, which we mostly have been for the last few years. Ocean temp is currently 7C (46F) which is a bit high, I think. But the ocean is quite variable around here, and also significantly moderates Boston at all times.

  23. The current slowing of AMOC appears to be a factor in the current jumps in both CO2 and CH4 atm. abundances. I estimate AMOC is responsible for ~5-8% of the total oceanic sink due to the increase of low-concentration water provided to the surface via overturning.

    However, at this moment, we have open water at the Bearing Straight. PIOMAS ice volume is being analyzed for the December average but it appears to be coming in around -8.5% below the previous record low.

    As the NH Summer insolation returns to the upper latitudes, this lack of ice will greatly amplify regional albedo impacts. This effect is going to rapidly increase as China continues to reduce its SO2 emissions and sea ice volume drives further ice declines. I estimate the first September ice free condition sometime around 2021 +/- 2 years and no later than 2025.

    Due to the effect of returning solar insolation and retreating ice edge boundaries, this increased solar absorption is very great. A 2013 Caldeira paper showed ~3.0 W/m^2 globally averaged forcing for a year-around ice free state. I estimate that an ice free condition of June 21 will be equal to the total forcing produced by a doubling of CO2 WITH LR/WV feedbacks. (~2.3 W/m^2).

    However, this effect will NOT be globally averaged but will instead be directed wholly at the Arctic Ocean. I expect this to occur sometime around 2065 but will very likely happen much more rapidly if recent studies of non-linearity of ECS are correct and ECS is closer to 4.5 vs. current 3.0.

    In this new regime, with a complete absence of sea ice and snow in the Northern Hemisphere, with rapid warming of the arctic region due to increased solar absorption, a jump in regional temps will occur. This rapid warming impact will greatly increase the meridional transport of 1000Hpa tropical water vapor into the region, leading to rainfall on the slopes of Greenland up to 600 meters elevation.

    At this point we will begin to see the Greenland pulse, on a scale of the Agassiz event. Though the dynamics described above may be sufficient to halt AMOC flow before that.

  24. Observed ice sheet mass loss doubling rates, although records are short, are ∼ 10 years (Sect. 5.1). Our sharp cutoff of melt aids separation of immediate forcing effects and feedbacks. We argue that such a rapid increase in meltwater is plausible if GHGs keep growing rapidly. Greenland and Antarctica have outlet glaciers in canyons with bedrock below sea level well back into the ice sheet (Fretwell et al., 2013; Morlighem et al., 2014; Pollard et al., 2015).

    Feedbacks, including ice sheet darkening due to surface melt (Hansen et al., 2007b; Robinson et al., 2012; Tedesco et al., 2013; Box et al., 2012) and lowering and thus warming of the near-coastal ice sheet surface, make increasing ice melt likely. Paleoclimate data reveal sea level rise of several meters in a century (Fairbanks, 1989; Deschamps et al., 2012). Those cases involved ice sheets at lower latitudes, but 21st century climate forcing is larger and increasing much more rapidly.

    Temperature change in 2065, 2080 and 2096 for 10-year doubling time (Fig. 6) should be thought of as results when sea level rise reaches 0.6, 1.7 and 5 m, because the dates depend on initial freshwater flux. Actual current freshwater flux may be about a factor of 4 higher than assumed in these initial runs, as we will discuss, and thus effects may occur ∼ 20 years earlier.

    A sea level rise of 5 m in a century is about the most extreme in the paleo-record (Fairbanks, 1989; Deschamps et al., 2012), but the assumed 21st century climate forcing is also more rapidly growing than any known natural forcing. Meltwater injected into the North Atlantic has larger initial impact, but Southern Hemisphere ice melt has a greater global effect for larger melt as the effectiveness of more meltwater in the North Atlantic begins to decline. The global effect is large long before sea level rise of 5 m is reached.

    Meltwater reduces global warming about half by the time sea level rise reaches 1.7 m. Cooling due to ice melt more than eliminates A1B warming in large areas of the globe. The large cooling effect of ice melt does not decrease much as the ice melting rate varies between doubling times of 5, 10 or 20 years (Fig. 7a). In other words, the cumulative ice sheet melt, rather than the rate of ice melt, largely determines the climate impact for the range of melt rates covered by 5-, 10- and 20-year doubling times.

    Thus if ice sheet loss occurs even to an extent of 1.7 m sea level rise (Fig. 7b), a large impact on climate and climate change is predicted. Greater global cooling occurs for freshwater injected into the Southern Ocean, but the cooling lasts much longer for North Atlantic injection (Fig. 7a). That persistent cooling, mainly at Northern Hemisphere middle and high latitudes (Fig. S7), is a consequence of the sensitivity, hysteresis effects, and long recovery time of the AMOC (Stocker and Wright, 1991; Rahmstorf, 1995, and earlier studies referenced therein).

    AMOC changes are described below. When freshwater injection in the Southern Ocean is halted, global temperature jumps back within two decades to the value it would have had without any freshwater addition (Fig. 7a). Quick recovery is consistent with the Southern Ocean-centric picture of the global overturning circulation (Fig. 4; Talley, 2013), as the Southern Ocean meridional overturning circulation (SMOC), driven by AABW formation, responds to change in the vertical stability of the ocean column near Antarctica (Sect. 3.7) and the ocean mixed layer and sea ice have limited thermal inertia.

    Cooling from ice melt is largely regional, temporary, and does not alleviate concerns about global warming. Southern Hemisphere cooling is mainly in uninhabited regions. Northern Hemisphere cooling increases temperature gradients that will drive stronger storms (Sect. 3.9).

    Global cooling due to ice melt causes a large increase in Earth’s energy imbalance (Fig. 7b), adding about +2 W m−2 , which is larger than the imbalance caused by increasing GHGs. Thus, although the cold freshwater from ice sheet disintegration provides a negative feedback on regional and global surface temperature, it increases the planet’s energy imbalance, thus providing more energy for ice melt (Hansen, 2005).

    This added energy is pumped into the ocean. Increased downward energy flux at the top of the atmosphere is not located in the regions cooled by ice melt. However, those regions suffer a large reduction of net incoming energy (Fig. 8a). The regional energy reduction is a consequence of increased cloud cover (Fig. 8b) in response to the colder ocean surface.

    However, the colder ocean surface reduces upward radiative, sensible and latent heat fluxes, thus causing a large (∼ 50 W m−2 ) increase in energy into the North Atlantic and a substantial but smaller flux into the Southern Ocean (Fig. 8c). Below we conclude that the principal mechanism by which this ocean heat increases ice melt is via its effect on ice shelves. Discussion requires examination of how the freshwater injections alter the ocean circulation and internal ocean temperature.

    3.5
    Simulated Atlantic meridional overturning circulation (AMOC)

    Broecker’s articulation of likely effects of freshwater outbursts in the North Atlantic on ocean circulation and global climate (Broecker, 1990; Broecker et al., 1990) spurred quantitative studies with idealized ocean models (Stocker and Wright, 1991) and global atmosphere–ocean models (Manabe and Stouffer, 1995; Rahmstorf 1995, 1996). Scores of modeling studies have since been carried out, many reviewed by Barreiro et al. (2008), and observing systems are being developed to monitor modern changes in the AMOC (Carton and Hakkinen, 2011).

    Our climate simulations in this section are five-member ensembles of runs initiated at 25-year intervals at years 901– 1001 of the control run. We chose this part of the control run because the planet is then in energy balance (Fig. S1), although by that time model drift had altered the slow deepocean circulation.

    Some model drift away from initial climatological conditions is inevitable, as all models are imperfect, and we carry out the experiments with cognizance of model limitations. However, there is strong incentive to seek basic improvements in representation of physical processes to reduce drift in future versions of the model. GHGs alone (scenario A1B) slow AMOC by the early 21st century (Fig. 9), but variability among individual runs (Fig. S8) would make definitive detection difficult at present. Freshwater injected into the North Atlantic or in both hemispheres shuts down the AMOC (Fig. 9, right side).

    GHG amounts are fixed after 2100 and ice melt is zero, but after two centuries of stable climate forcing the AMOC has not recovered to its earlier state. This slow recovery was found in the earliest simulations by Manabe and Stouffer (1994) and Rahmstorf (1995, 1996). Freshwater injection already has a large impact when ice melt is a fraction of 1 m of sea level.

    By the time sea level rise reaches 59 cm (2065 in the present scenarios), when freshwater flux is 0.48 Sv, the impact on AMOC is already large, consistent with the substantial surface cooling in the North Atlantic (Fig. 6). 3.6 Comparison with prior simulations AMOC sensitivity to GHG forcing has been examined extensively based on CMIP studies.

    Schmittner et al. (2005) found that AMOC weakened 25 ± 25 % by the end of the 21st century in 28 simulations of 9 different models forced by the A1B emission scenario. Gregory et al. (2005) found 10– 50 % AMOC weakening in 11 models for CO2 quadrupling (1 % year−1 increase for 140 years), with largest decreases in models with strong AMOCs. Weaver et al. (2007) found a 15–31 % AMOC weakening for CO2 quadrupling in a single model for 17 climate states differing in initial GHG amount.

    AMOC in our model weakens 30 % in the century between 1990–2000 and 2090–2100, the period used by Schmittner et al. (2005), for A1B forcing (Fig. S8). Thus our model is more sensitive than the average but within the range of other models, a conclusion that continues to be valid in comparison with 10 CMIP5 models (Cheng et al., 2013). AMOC sensitivity to freshwater forcing has not been compared as systematically among models.

    Several studies find little impact of Greenland melt on AMOC (Huybrechts et al., 2002; Jungclaus et al., 2006; Vizcaino et al., 2008) while others find substantial North Atlantic cooling (Fichefet et al., 2003; Swingedouw et al., 2007; Hu et al., 2009, 2011). Studies with little impact calculated or assumed small ice sheet melt rates, e.g., Greenland contributed only 4 cm of sea level rise in the 21st century in the ice sheet model of Huybrechts et al. (2002). Fichefet et al. (2003), using nearly the same atmosphere–ocean model as Huybrechts et al. (2002) but a more responsive ice sheet model, found AMOC weakening from 20 to 13 Sv late in the 21st century, but separate contributions of ice melt and GHGs to AMOC slowdown were not defined. Hu et al. (2009, 2011) use the A1B scenario and freshwater from Greenland starting at 1 mm sea level per year increasing 7 % year−1 , similar to our 10-year doubling case.

    Hu et al. keep the melt rate constant after it reaches 0.3 Sv (in 2050), yielding 1.65 m sea level rise in 2100 and 4.2 m in 2200. Global warming found by Hu et al. for scenario A1B resembles our result but is 20–30 % smaller (compare Fig. 2b of Hu et al., 2009 to our Fig. 6), and cooling they obtain from the freshwater flux is moderately less than that in our model. AMOC is slowed about one-third by the latter 21st century in the Hu et al. (2011) 7 % year−1 experiment, comparable to our result. General consistency holds for other quantities, such as changes of precipitation.

    Our model yields southward shifting of the Intertropical Convergence Zone (ITCZ) and intensification of the subtropical dry region with increasing GHGs (Fig. S9), as has been reported in modeling studies of Swingedouw et al. (2007, 2009). These effects are intensified by ice melt and cooling in the North Atlantic region (Fig. S9). A recent five-model study (Swingedouw et al., 2014) finds a small effect on AMOC for 0.1 Sv Greenland freshwater flux added in 2050 to simulations with a strong GHG forcing. Our larger response is likely due, at least in part, to our freshwater flux reaching several tenths of a sverdrup.

    3.7
    Pure freshwater experiments
    We assumed, in discussing the relevance of these experiments to Eemian climate, that effects of freshwater injection dominate over changing GHG amount, as seems likely because of the large freshwater effect on sea surface temperatures (SSTs) and sea level pressure.

    However, Eemian CO2 was actually almost constant at ∼ 275 ppm (Luthi et al., 2008). Thus, to isolate effects better, we now carry out simulations with fixed GHG amount, which helps clarify important feedback processes. Our pure freshwater experiments are five-member ensembles starting at years 1001, 1101, 1201, 1301, and 1401 of the control run. Each experiment ran 300 years.

    Freshwater flux in the initial decade averaged 180 km3 year−1 (0.5 mm sea level) in the hemisphere with ice melt and increased with a 10-year doubling time. Freshwater input is terminated when it reaches 0.5 m sea level rise per hemisphere for three five member ensembles: two ensembles with injection in the individual hemispheres and one ensemble with input in both hemispheres (1 m total sea level rise). Three additional ensembles were obtained by continuing freshwater injection until hemispheric sea level contributions reached 2.5 m.

    Here we provide a few model diagnostics central to discussions that follow. Additional results are provided in Figs. S10–S12. The AMOC shuts down for Northern Hemisphere freshwater input yielding 2.5 m sea level rise (Fig. 10). By year 300, more than 200 years after cessation of all freshwater input, AMOC is still far from full recovery for this large freshwater input. On the other hand, freshwater input of 0.5 m does not cause full shutdown, and AMOC recovery occurs in less than a century.

    Global temperature change (Fig. 11) reflects the fundamentally different impact of freshwater forcings of 0.5 and 2.5 m. The response also differs greatly depending on the hemisphere of the freshwater input. The case with freshwater forcing in both hemispheres is shown only in the Supplement because, to a good approximation, the response is simply the sum of the responses to the individual hemispheric forcings (see Figs. S10–S12).

    The sum of responses to hemispheric forcings moderately exceeds the response to global forcing. Global cooling continues for centuries for the case with freshwater forcing sufficient to shut down the AMOC (Fig. 11). If the forcing is only 0.5 m of sea level, the temperature recovers in a few decades.

    However, the freshwater forcing required to reach the tipping point of AMOC shutdown may be less in the real world than in our model, as discussed below. Global cooling due to freshwater input on the Southern Ocean disappears in a few years after freshwater input ceases (Fig. 11), for both the smaller (0.5 m of sea level) and larger (2.5 m) freshwater forcings.

    Injection of a large amount of surface freshwater in either hemisphere has a notable impact on heat uptake by the ocean and the internal ocean heat distribution (Fig. 12). Despite continuous injection of a large amount of very cold (−15 ◦C) water in these pure freshwater experiments, substantial portions of the ocean interior become warmer.

    Tropical and Southern Hemisphere warming is the well-known effect of reduced heat transport to northern latitudes in response to the AMOC shutdown (Rahmstorf, 1996; Barreiro et al., 2008). However, deep warming in the Southern Ocean may have greater consequences. Warming is maximum at grounding line depths (∼ 1–2 km) of Antarctic ice shelves (Rignot and Jacobs, 2002).

    Ice shelves near their grounding lines (Fig. 13 of Jenkins and Doake, 1991) are sensitive to temperature of the proximate ocean, with ice shelf melting increasing 1 m per year for each 0.1 ◦C temperature increase (Rignot and Jacobs, 2002). The foot of an ice shelf provides most of the restraining force that ice shelves exert on landward ice (Fig. 14 of Jenkins and Doake, 1991), making ice near the grounding line the buttress of the buttress.

    Pritchard et al. (2012) deduce from satellite altimetry that ice shelf melt has primary control of Antarctic ice sheet mass loss. Thus we examine our simulations in more detail (Fig. 13). The pure freshwater experiments add 5 mm sea level in the first decade (requiring an initial 0.346 mm year−1 for 10-year doubling), 10 mm in the second decade, and so on (Fig. 13a). Cumulative freshwater injection reaches 0.5 m in year 68 and 2.5 m in year 90.

    Antarctic Bottom Water (AABW) formation is reduced ∼ 20 % by year 68 and ∼ 50 % by year 90 (Fig. 13b). When freshwater injection ceases, AABW formation rapidly regains full strength, in contrast to the long delay in reestablishing North Atlantic Deep Water (NADW) formation after AMOC shutdown.

    The Southern Ocean mixed-layer response time dictates the recovery time for AABW formation. Thus rapid recovery also applies to ocean temperature at depths of ice shelf grounding lines (Fig. 13c).

    The rapid response of the Southern Ocean meridional overturning circulation (SMOC) implies that the rate of freshwater addition to the mixed layer is the driving factor. Freshwater flux has little effect on simulated Northern Hemisphere sea ice until the 7th decade of freshwater growth (Fig. 13d), but Southern Hemisphere sea ice is more sensitive, with substantial response in the 5th decade and large response in the 6th decade. Below we show that “5th decade” freshwater flux (2880 Gt year−1 ) is already relevant to the Southern Ocean today.

    […]

    3.9
    Impact of ice melt on storms
    Our inferences about potential storm changes from continued high growth of atmospheric GHGs are fundamentally different than modeling results described in IPCC (2013, 2014), where the latter are based on CMIP5 climate model results without substantial ice sheet melt. Lehmann et al. (2014) note ambiguous results for storm changes from prior model studies and describe implications of the CMIP5 ensemble of coupled climate models.

    Storm changes are moderate in nature, with even a weakening of storms in some locations and seasons. This is not surprising, because warming is greater at high latitudes, reducing meridional temperature gradients. Before describing our model results, we note the model limitations for study of storms, including its coarse resolution (4◦ × 5 ◦ ), which may contribute to slight misplacement of the Bermuda high-pressure system for today’s climate (Fig. S2).

    Excessive Northern Hemisphere sea ice may cause a bias in location of deepwater formation toward lower latitudes. Simulated effects also depend on the location chosen for freshwater injection; in model results shown here (Fig. 20), freshwater was spread uniformly over all longitudes in the North Atlantic between 65◦ W and 15◦ E. It would be useful to carry out similar studies with higher resolution models including the most realistic possible distribution of meltwater.

    Despite these caveats, we have shown that the model realistically simulates meridional changes of sea level pressure in response to climate forcings (Sect. 3.8.5). Specifically, the model yields a realistic trend to the positive phase of the Southern Annular Mode (SAM) in response to a decrease in stratospheric ozone and increase in other GHGs (Fig. S17). We also note that the modeled response of atmospheric pressure to the cooling effect of ice melt is large scale, tending to be of a meridional nature that should be handled by our model resolution.

    Today’s climate, not Eemian climate, is the base climate state upon which we inject polar freshwater. However, the simulated climate effects of the freshwater are so large that they should also be relevant to freshwater injection in the Eemian period.

    3.9.1
    Modeling insights into Eemian storms
    Ice melt in the North Atlantic increases simulated sea level pressure in that region in all seasons (Fig. 20).

    In summer the Bermuda high-pressure system (Fig. S2) increases in strength and moves northward. Circulation around the high pressure creates stronger prevailing northeasterly winds at latitudes of Bermuda and the Bahamas. A1B climate forcing alone (Fig. S21, top row) has only a small impact on the winds, but cold meltwater in the North Atlantic causes a strengthening and poleward shift of the high pressure.

    The high pressure in the model is located further east than needed to produce the fastest possible winds at the Bahamas. Our coarse-resolution (4◦ × 5 ◦ ) model may be partly responsible for the displacement. However, the location of high pressure also depends on meltwater placement, which we spread uniformly over all longitudes in the North Atlantic between 65◦ W and 15◦ E, and on the specific location of ocean currents and surface temperature during the Eemian.

    North Atlantic cooling from AMOC shutdown creates faster winds in our simulations, with a seasonal-mean increment as much as 10–20 %. Such a percentage translates into an increase in storm power dissipation by a factor ∼ 1.4–2, because dissipation is proportional to the cube of wind speed (Emanuel, 1987, 2005).

    Our simulated changes refer to mean winds over large grid boxes, not individual storms, for which the change in the most extreme cases might be larger. Increased North Atlantic high pressure strengthens prevailing northeasterly winds blowing onto the Bahamas in the direction of Eemian wave-formed deposits (Sect. 4.1.2).

    Consistent increase in these winds would contribute to creation of long-wavelength, deep-ocean waves that scour the ocean floor as they reach the shallow near-shore region. However, extreme events may require the combined effect of increased prevailing winds and tropical storms guided by the strengthened blocking high pressure and nurtured by the unusually warm late-Eemian tropical sea surface temperatures (Cortijo et al., 1999), which would favor more powerful tropical storms (Emanuel, 1987).

    This enhanced meridional temperature gradient – warmer tropics and cooler high latitudes – was enhanced by low obliquity of Earth’s spin axis in the late Eemian.

    3.9.2
    21st century storms
    If GHGs continue to increase rapidly and ice melt grows, our simulations yield shutdown or major slowdown of the AMOC in the 21st century, implying an increase in severe weather. This is shown by zonal-mean temperature and eddy kinetic energy changes in simulations of Sects. 3.3–3.6 with and without ice melt (Fig. 21).

    Without ice melt, surface warming is largest in the Arctic (Fig. 21, left), resulting in a decrease in lower tropospheric eddy energy. However, the surface cooling from ice melt increases surface and lower tropospheric temperature gradients, and in stark contrast to the case without ice melt, there is a large increase in midlatitude eddy energy throughout the midlatitude troposphere.

    The increase in zonal-mean midlatitude baroclinicity (Fig. 21) is in agreement with the localized, North Atlantic-centered increases in baroclinicity found in the higher-resolution simulations of Jackson et al. (2015) and Brayshaw et al. (2009). Increased baroclinicity produced by a stronger temperature gradient provides energy for more severe weather events. Many of the most significant and devastating storms in eastern North America and western Europe, popularly known as superstorms, have been winter cyclonic storms, though sometimes occurring in late fall or early spring, that generate near-hurricane-force winds and often large amounts of snowfall (Chapter 11, Hansen, 2009).

    Continued warming of low-latitude oceans in coming decades will provide a larger water vapor repository that can strengthen such storms. If this tropical warming is combined with a cooler North Atlantic Ocean from AMOC slowdown and an increase in midlatitude eddy energy (Fig. 21), we can anticipate more severe baroclinic storms. Increased high pressure due to cooler high-latitude ocean (Fig. 20) can make blocking situations more extreme, with a steeper pressure gradient between the storm’s low-pressure center and the blocking high, thus driving stronger North Atlantic storms.

    Freshwater injection into the North Atlantic and Southern oceans increases sea level pressure at middle latitudes and decreases it at polar latitudes (Figs. 20, S22), but the impact is different in the North Atlantic than in the Southern Ocean.

    In the Southern Ocean the increased meridional temperature gradient increases the strength of westerlies in all seasons at all longitudes. In the North Atlantic Ocean the increase in sea level pressure in winter slows the westerlies (Fig. 20). Thus instead of a strong zonal wind that keeps cold polar air locked in the Arctic, there is a tendency for a less zonal flow and thus more cold air outbreaks to middle latitudes.

    http://csas.ei.columbia.edu/2015/09/21/predictions-implicit-in-ice-melt-paper-and-global-implications/

  25. Stefan wrote:
    “The decisive factor is whether the AMOC brings freshwater into the Atlantic basin or whether it exports it (in the latter case, working to increase salinity in the Atlantic). My article ended with the suggestion to clarify this from observational data. That was later done by colleagues from Holland (Weijer et al. 1999).”

    The answer was not clarified from observational data. Observational data was fed into a 2D model, and the model returned an answer based upon how the model was constructed. To clarify with observational data, you would need to directly measure the fluxes.

  26. #6 Joe Neubarth

    “With CO2 doubling and Methane being added by leaps and bounds and Nitrous Oxide just starting a massive ramp-up, I doubt that these projections can be valid beyond the short term.”

    Well, we have to start somewhere, don’t we? What other scenarios would you suggest, and in which direction – climatically speaking – should they be aimed? Are you suggesting a much faster rate of warming should be used in the climate forcing? Presumably much faster warming would generate much faster melt of Greenland and other Arctic ice fields, and therefore an even more severe shock to the (AMOC) system.
    +++++++++++++++++++++++++

    “I see no mentioning of the fact that way over 90 percent of the Arctic Sea Ice will not be there to melt and produce large amounts of freshwater content. The Greenland melt should continue, but as the ice melts in coverage area (more land is exposed) the melting ice run off should decrease.”

    Assuming that this is indeed ‘fact’, there will still be input from precipitation into the sea, and direct runoff from exposed land areas which are no longer ice covered e.g. Greenland, Baffin Island and so on. There’s also the possibility of increased fluvial input from the major rivers flowing into the Arctic Ocean; not sure what the climate models indicate in this respect, someone else here may know. Possibly the issue to query here is what happens if the AMOC shuts down/slows down significantly. Specifically, what would it take for the system to start working again, at the current rate or something approaching what we are familiar with?
    ++++++++++++++++++++

    “Nobody mentioned the stability of the salt content of the oceans, but with all that water vapor going into the sky I would expect salt content to go up. Is that written into the projections for 2100?”

    It is all a matter of mass balance, and where it is distributed spatially. I very much doubt that all the melted ice will end up in the atmosphere as water vapour; be interesting to see if someone has some figures on this. As regards a moister atmosphere in a warmer world, my understanding is that this will lead to more precipitation in the polar and temperate latitudes. All that extra evaporation has to go somewhere, and some areas will probably get much wetter as a result. We had an example of this in the U.K. last Winter, 2015/2016, following the v. strong El Nino, and the atmospheric river effect which brought record rainfall and flooding to much of the British Isles. We also had a warm and wet winter in 2013/2014. Some climate models predict El Nino occurring with greater frequency – almost annually? – and magnitude, which implies a lot more of the same. Difficult to see how salt content of the polar and temperate ocean waters is going to go up in those circumstances.
    ++++++++++++++++++++

  27. AMOC carries oxygen-rich water to the ocean depths, thousands of metres from the atmosphere, where it sustains complex organisms. Do we have any idea what the consequences are of the kind of slowdown of AMOC suggested by this latest study in terms of ocean anoxia and euxinia. Are we looking at a future of oceans, and eventually the atmosphere, suffused with hydrogen sulphide?

  28. Hank @8,

    Did you notice that the NCAR diagram from Yeager et al (2015) shows that their model underestimates the melting of the sea ice up until 2007?
    https://www2.ucar.edu/sites/default/files/news/2015/Sea%20Ice%20FINAL-01%20Lat%20Long1500.jpg

    The sea ice retreated even further in 2012. And in 2016, the winter Arctic sea ice extent has been running at a new all time low for the time of year for several months now, which you can verify here if you select 2016. http://nsidc.org/arcticseaicenews/charctic-interactive-sea-ice-graph/

    How can we accept their prediction that the Arctic sea ice will show little increase in melt over the next decade? It seems to me more likely that the inability to fully recover during this winter will lead to more loss during the summer making it yet more difficult for the sea ice to refreeze during the winter. In other words it has passed a tipping point into a positive feedback loop where less summer ice means less winter ice anc less winter ice leads to less summer ice, leading to it melt away abruptly and completely!

  29. Correct link to above Hansen et al 2016 study quote on AMOC implications and so on.
    http://www.atmos-chem-phys.net/16/3761/2016/acp-16-3761-2016.pdf

    The main take away is that a AMOC shutdown is a likely scenario under BAU, and possible implications include a dark scenario of irreversible altercation of our environment, effecting Northern Hemisphere lives, and pose undiscussed new logistic and travel challenges, due to stronger winds. This means how safe will ship and air travel be for intercontinental exchange?

    If we go a little farther we can expect geomorphological hazards due to mass shifts, waves, winds, rain patterns, as outlined in this video https://www.youtube.com/watch?v=xndhx7KpSU0

  30. The answer was not clarified from observational data. Observational data was fed into a 2D model, and the model returned an answer based upon how the model was constructed. To clarify with observational data, you would need to directly measure the fluxes.

    I suppose there is something to be said for maintaining the traditional distinction between modeled and strictly ‘observational’ data. But as Paul Edwards documents at considerable length in his “A Vast Machine”, it’s a distinction that seems a bit quaint when the ever-growing reliability of weather forecasts is founded overwhelmingly on modeled, not directly observational, data.

    After all, people do customarily construct those models to give accurate results. That’s kind of the point.

    [Response: There is almost no such thing as pure observed data. Already reading temperature off a mercury thermometer requires a model of how the length of the mercury column correlates with temperature. -stefan]

  32. Slioch @27 — Have you read “Six Degrees” by Mark Lynas? That is likely to answer your question.

  33. 24 prokaryotes … 5 stars :-)

    27 Slioch says: “Are we looking at a future of oceans, and eventually the atmosphere, suffused with hydrogen sulphide?”

    Paleoclimate science and earth history suggests that is entirely possible. But of course “it depends” solely on the present…. As each day and year unfolds it’s always the present, so it always “depends”.

    Prof. Peter Ward UW put the ‘yardstick’ as being if/when the Earth hits 1000ppm of CO2 it’s almost guaranteed there will be no ice at the poles and the hydrogen sulphide world returns for millions of years yet again.

    In my mind it kinda works like 1+1 = 2 and beyond self-evident. Still people don’t care nor believe in climate science nor scientists and especially not the environmentalists. Kinda a catch 22

  34. Kevin McKinney wrote:
    “But as Paul Edwards documents at considerable length in his “A Vast Machine”, it’s a distinction that seems a bit quaint when the ever-growing reliability of weather forecasts is founded overwhelmingly on modeled, not directly observational, data.”

    These are important concepts, so it is useful to understand them. Weather is indeed predicted with models. However, the predictand, the actual weather, is directly measured. Therefore the fidelity of the weather models can be assigned actual numbers, which we call skill. Climate models predict future climate in the same way, but the fidelity/skill cannot be directly measured because we lack observational data about future climate. There is a difference.

    Stefan obviously did not attempt to mislead. But when he wrote that the question of net flux was clarified by observational data, he inadvertently implied that the fidelity of his model was verified by observational data in the same way that a weather prediction is validated with thermometer measurements, and that is not what happened. It is an important distinction.

    Stefan wrote:
    “Already reading temperature off a mercury thermometer requires a model of how the length of the mercury column correlates with temperature.”

    I do not agree. Let’s step through this. I have a little vial of mercury, and on a particularly hot day, the cork pops out. Hmmm, methinks. So I construct a thin glass tube with a bulb of mercury at the bottom. I put it in ice water, and I scribe the glass at the mercury line. I then put it in boiling water, notice that the mercury is higher in the glass, and I make a second scribe. I use my ruler to put 98 more scribes in between.

    I don’t need to know why or how the mercury expands, or whether it does so linearly. I did not model anything. There is no model to feed data into. But I do have a pretty damn useful way of measuring temperature. If I wanted to answer the question of whether a glass tube that is twice the diameter will have the scribes the same distance apart, I have two choices, I can measure it empirically or I can model it based upon some understanding about the expansion of fluids. It is not all modeling.

    [Response: But your marks have no connection to anything else until it’s calibrated – which requires some kind of model. It’s the same way that a proxy measurement of temperature is used. – gavin]

  35. 33
    Thomas says:
    5 Jan 2017 at 11:20 PM

    24 prokaryotes … 5 stars :-)

    Guys, there’s this wonderful new Internet thing called hyperlinks, very much worth using for long articles instead of pasting the whole text into a comment. Just give us a brief summary/remark on why you think it’s worth reading, and provide a hyperlink.

  36. #30 re “[Response: There is almost no such thing as pure observed data. Already reading temperature off a mercury thermometer requires a model of how the length of the mercury column correlates with temperature. -stefan]”

    That was an unexpected and curious thing to say on the subject. Whilst fundamentally true and correct, obviously, it is still quite odd to me.

    Like burrowing down into that proverbial rabbit hole, it’s not always helpful in the long term. :-)

  37. Thomas, #33
    “Prof. Peter Ward UW put the ‘yardstick’ as being if/when the Earth hits 1000ppm of CO2 it’s almost guaranteed there will be no ice at the poles and the hydrogen sulphide world returns for millions of years yet again”.

    This is the part that interests / concerns me: “and the hydrogen sulphide world returns for millions of years yet again”.
    Could you give me the reference, please? I’m a geologist. Anoxic environments and temperature anomalies (high and low) are of professional interest.

  38. Would this be of any help?

    Overlooked possibility of a collapsed Atlantic Meridional Overturning Circulation in warming climate

    Wei Liu1,*,†, Shang-Ping Xie1, Zhengyu Liu2 and Jiang Zhu2
    http://advances.sciencemag.org/content/3/1/e1601666.full

    By correcting the model biases, we show that the AMOC collapses 300 years after the atmospheric CO2 concentration is abruptly doubled from the 1990 level.

  39. #31,#32,#33 Thanks for your responses, particularly the RS lecture series, which I will study.
    I have to agree with Thomas: a hydrogen sulphide world is looking, to me, almost inevitable with bau. Peter Ward’s warnings are eloquent. His “we do not know of a time with permanent ice at the poles and CO2 above 1000pmmv” (except, of course, prior to the big thaw in snowball Earth), and the present rate of increase of atmospheric CO2 being c.10x greater than previous mass extinctions as far as we know (albeit the total mass being less) are deeply worrying. It is one thing to raise sea levels by several metres, ravage climates, devastate human civilisation and wipe out half of global biodiversity, but a hydrogen sulphide world is an order of magnitude more horrific. And yet hardly anyone seems to talk about it.
    I was rather hoping someone would come along and assure me that the dangers of a breakdown of the Gulf Stream System did not include such horrors.

  40. If the temperatures in the north atlantic region would drop so much, the melting of the greenland ice shield might be stopped completely – which would be a positive aspect of the whole thing. Is this true?

  41. 37 Richard Hawes hi, you’d be better off contact Ward direct, he’ll reply to you because he replied to me even when working on a dig site when in South Australia. https://www.ess.washington.edu/people/profile.php?pid=ward–peter

    best example paleoclimate ref on youtube is this: https://www.youtube.com/watch?v=HP_Fvs48hb4&feature=youtu.be&t=33m1s

    Google scholar has his and colleagues papers, just search https://scholar.google.com.au/scholar?q=%22pd+ward%22+&btnG=&hl=en&as_sdt=0%2C5

  42. #34 It’s the same way that a proxy measurement of temperature is used. – gavin

    Sure, but personally I would use the word “similar” and not “same”; and the word “process” and not “way”. Words matter mate. imho there’s a huge difference between the direct accuracy of a marked thermometer versus tree rings and ice cores etc.

    Gavin, you have said often “models are useful” and I totally agree. However when it comes to a mercury thermometer or a new digital one they are more than useful, they 99.9% accurate as far as determining a present “observation”. eg argo floats today versus GCMs modeled SSTs in the past.

    Beware the semantic and jargon traps Gavin, that’s all I’m saying (not complaining). Thermometers are not equivalent to other “models” simply because modelling is used to put marks on a thermometer. The way you and Stefan have said it muddies the waters for the average person in the public space, imho.

    PS please find an excuse to come to Australia in 2017 and get a gig on Q&A. Most aussies would warmly welcome you here speaking about what you know, unlike the US Congress. ;-)

  43. ps Gavin, you have said often “models are useful” and I totally agree .. and also said “all models are wrong” too.

    I shouldn’t trust the nurse taking my temperature is not too high because the model it is based on is ‘wrong’?

    Analogies are useful, and yet must be used with caution and clarity ….. :-)

  44. I think the only question is whether the shutdown will happen within the next few weeks or several months from now.

    The rapid decline in the sea ice combined with all the fresh water coming off Greenland has to mean it is going to happen soon.

  46. Matt Skaggs, #34:

    “Weather is indeed predicted with models. However, the predictand, the actual weather, is directly measured.”

    Yes, but much less than you think, or than I thought before reading “A Vast Machine.” According to Edwards, only about 10% of the data comprising a weather “analysis”–that’s the term for the data model which is the numerical “picture” of the atmospheric state preceding the forecast–is derived from observation. The rest comes from the preceding forecast, which serves as “first guess” for the new analysis. As he puts it, the observational data “constrains but does not determine”.

    It sounds a bit crazy, coming from a “traditional” perspective. But there are good reasons for that reality. One is that the 10% that’s observational is already a huge mass of data, obtained with considerable difficulty and expense; all the additional observation required for purely observational analyses would be utterly impractical on multiple counts. (I suspect that the same would be true of the flux measurements mentioned upthread.)

    Another is that it would still have to be modeled anyway, because observation is not perfect; an important part of preparing an analysis has always been data control, including the recognition and correction or discarding of spurious readings.

    A third is that observations do not conform to model structure, which demands a homogenous data field. Observations cannot be taken at each and every model gridpoint. So you can interpolate the value for the gridpoint–a process implicit in traditional ‘manual’ weather analysis–or you can take the modeled value from the previous cycle, which was calculated on simulated physical processes at least–a better alternative to interpolation/extrapolation. A corollary of this is that you end up comparing data model to forecast, not raw observation to forecast.

    The ultimate justification of this is that it works. Not perfectly, of course; Atlanta for instance just got shut down on the basis of forecast snow that turned out to seriously overpredict accumulations. (Though Asheville got about 5 inches, and I’m hearing the mid-Atlantic coast may be hit harder.) But it works a lot better than forecasting used to, as I well recall, and as objective evaluation of skill demonstrates.

    A dramatic example from “AVM” is found in Figure 10.8, p. 273. Way back on August 29, 1985, the European Met analysis showed a synoptic scale eddy in the western Sahara, hundreds of kilometers from any station reporting data. (There were stations closer than that, but the regional telecom net had gone down on the 25th, so none of the observations got through.) Yet the vortex was clearly visible on satellite imagery, just as calculated on the basis of 100% modeled data.

  47. If we were to use “Ocean Mechanical Thermal Energy Conversion” “OMTEC” to our advantage this will never happen folks.

  48. @39,
    A very good overview of the oxygen depletion issue is here:
    http://news.nationalgeographic.com/2015/03/150313-oceans-marine-life-climate-change-acidification-oxygen-fish/

    The natural thing to expect is that as the ocean gets warmer, circulation will slow down and get more sluggish and the waters going into the deep ocean will hang around longer,” says Curtis Deutsch, a chemical oceanography professor at the University of Washington, in Seattle. “And indeed, oxygen seems to be declining.”

    The sources of oxygenated water to the deep ocean are the polar regions. North Atlantic Deep Water (NADW) forms as part of the AMOC;Antarctic Bottom Water (ABW) forms due to the Antarctic sea ice system, and Antarctic Intermediate Water (AAIW) forms at the Antarctic Polar Front (50-60 S). All three play roles in oxygenating the deep ocean (those are wiki links with lots of good information); so you have to consider all three in any forecast of oxygen ventilation changes.

    It’s not just the circulation, however; waming ocean waters mean higher rates of microbial decomposition of sinking organic matter, which sucks oxygen out of the water. The huge amounts of agricultural fertilizers dumped into rivers and coastal waters are also playing a role, as in the Gulf of Mexico dead zone. So you have a whole host of complex factors, making it hard to forecast – along with a relative lack of data on historical deep ocean oxygen levels. It’s a serious concern, however, and is getting more attention:

    https://www.scientificamerican.com/article/ocean-s-oxygen-starts-running-low/

    https://www.washingtonpost.com/news/energy-environment/wp/2016/04/28/global-warming-could-deplete-the-oceans-oxygen-levels-with-severe-consequences/

    https://www2.ucar.edu/atmosnews/news/20721/widespread-loss-of-ocean-oxygen-become-noticeable-in-2030s

    So, as far as your question, under business-as-usual is it plausible to convert the whole ocean into something like the Black Sea, with only the upper 10% of the ocean having enough oxygen to support aerobic life? It would take a long time; but if you go and burn all the fossil fuels and hit 1000 ppm CO2, it doesn’t seem out of the question.

  49. Just an aesthetic observation, if the area in the graphic does become a whole lot colder then we would have an utterly weird northern hemisphere with the Canadian arctic islands and Alaska, northern Siberia and the Arctic ocean ice free and warmish, and Greenland, Iceland, Svalbard, Scandinavia and the North Atlantic frozen solid!

    Just bizarre.