top of page
  • Writer's pictureAllan Dyen-Shapiro

Meltwater and Overturning Circulation: Allan digests James Hansen's latest paper

A few weeks ago, I linked on my author Facebook profile (do friend me via the link at the bottom of this page) to the latest paper from James Hansen, NASA's top climate scientist from 1981 to 2013 and now a professor at Columbia University, and promised I'd digest it and blog on it. It was a definite slog; it took me a while. The core of the paper is mathematical modeling. I did mathematical modeling at one point in my science career, and for the uninitiated, I describe it as a way to know if you are telling the truth. If I cannot model a complex process, no matter what reasonable assumptions I make there are no solutions that match the real-world data, I can conclude that either the data is faulty, my thinking is faulty or both. The proper response is to check the data and/or think in different ways. However, once you have a model that passes this test, you can use it to ask "What if" questions. If this particular parameter changes, what happens? The key question Hansen asked is how sensitive is the global climate system to the increase in meltwater from glaciers in Antarctica and Greenland. More specifically, how important are certain positive feedback loops?

Positive feedback is where one thing leads to another, and that second thing causes the rate at which the original change happens to increase. It tends to give responses that occur more like an on-off switch on a light than like a dimmer switch. In 2008, when I first started teaching high school, the literature on climate change spoke of two doomsday scenarios based on positive feedback: (1) Global warming melts permafrost in Siberia, releasing methane, causing much more global warming; (2) Polar ice melts, exposing sea water, which is pretty close to black in terms of light absorbance as opposed to white, and black things absorb more heat, and this extra heat melts more ice. Well, both are now facts of our existence. And yes, this is doomsday, there is now no hope that the planet will get out of this crisis unscathed; the only question is how bad things will be.

And now, having been ignored by politicians at the time despite being absolutely correct on previous issues of feedback, Hansen has done the science to argue that there are new positive feedback loops to worry about. And the popular press does what it now does, reporting mostly the estimates of increased sea level rise and the call to do whatever we can right now to shut down all use of fossil fuels, because even the 2% warming goal of the recent global agreement, already thought to be way too much, will not even be close to attainable even with perfect compliance with this agreement. Sure, this is an important message, but the public is tuning it out. Yeah, hundreds of millions of people will be displaced and all economies will be wrecked, but what's the latest with Caitlyn Jenner?

If you've read this far, I am assuming you are more intelligent than much of the vast unwashed. (Stealing from Shakespeare there; it's one of my favorite of the Bard's phrases.) Here's what I didn't do for you: (1) I didn't master the entire field of climate physics and chemistry, including every important paper ever written, so I could fact check every one of his assumptions. The ballpark estimate is that it takes a scientist ten years of full time work until they can truly master a new field to that extent. And I have this thing called a job that would have gotten in the way; besides, the paper has only been out for a few weeks. When I published mathematical modeling, I can truly say I had the entire biological field in which I worked in my head, and I had actually sat down and looked at methods particulars of every single important paper in order to judge which data was compatible and which I had to re-do myself to get a consistent dataset and then did the experiments before publishing a model. On one of my rejected grant proposals, several experts (not the ones who recommended rejection) independently stated that I was one of only two or three people in the world who could justifiably make that claim for my sub-field. James Hansen is in this position for this paper. I am not and never will be. (2) I didn't download the code and repeat his experiments, trying new mathematical formalisms, coding strategies, and simulations he didn't think of in order to double check him. He did this not only for his own model but for many other models, mostly from the IPCC, hundreds of scientists who are all experts (and who won the Nobel Prize for their work). But, with how important this field is, I assure you many other scientists worldwide are doing just this. If Hansen slipped up, it will come out soon in the peer-reviewed literature. (3) I didn't even look at his code. Understanding the code would have taken me many months of full-time work including many phone calls to those more expert than I and poring through many math textbooks.

All I did was read the paper, come to a surface understanding of everything in the paper, and boil it down into terms I hope an educated and intelligent audience of mostly nonscientists can understand. And even that took me three consecutive evenings. Here goes:

Previous models hadn't included the thermohaline stratification along the coast of Antartica and Greenland. The surface water is cold; the further down you go, the warmer it is. Cold, salty water is dense; it sinks. Look in your refrigerator at your vinegar and oil salad dressing. The oil is on top. Oil is low density; water (vinegar) is higher density, so it is on the bottom. When glaciers melt, they inject cold freshwater near the surface. The freshwater dilutes the surface water, so the sinking you normally see in the global North happens less readily. If cold water isn't sinking, warm water isn't rising (because of being pushed out of the way). As such, heat is trapped in deep water in the Atlantic; this heat can further melt glaciers (positive feedback). In the Southern Ocean (everything South of sixty degrees latitude; the ocean that surrounds Antarctica), greater stratification prevents heat from venting from the bottom water into the atmosphere. Venting of heat from the Southern Ocean is well known to be of major climatic significance. Often this heat is through holes in the ice (called polynyas) and very few have been seen in recent years as a consequence of global climate change.

So, the key change in Hansen's model is the water column is modeled as discrete element of volume. Transport of heat (and carbon dioxide) from bottom water to atmosphere is thus coded in equations as sequential transfer from one layer to the next.

Paradoxically, local cooling of surface water due to meltwater and thus local cooling of the atmosphere near the glaciers leads to increased heat trapped at depths in the ocean, even as you see more sea ice due to the surface cooling. Got that? Totally counterintuitive, but hopefully I was able to step you through the logic.

And there's real world evidence this is already happening. A study of ventilation of heat from bottom water and deep water in the Weddell Sea (Antarctica) between 1984 and 2008 revealed a 15 - 21% reduction (this is the range of uncertainty) over that time period.

Effects related to the degree of stratification are one of two positive feedback loops newly considered in this paper. The other has to do with P - E (precipitation minus evaporation) in latitudes somewhat north of Antarctica. Global warming leads to warmer oceans leads to more evaporation, so without considering meltwater, you'd expect that some of that would be snowed back onto Antartica and increase the ice pack. A good thing. However, the surface cooling cools the atmosphere. As a result, the water is rained back down into the ocean before it ever gets to Antarctica. The more meltwater, the less new snow/ice in Antarctica. Again, positive feedback. Hansen incorporated current measured data on this issue into the new model.

The other key conclusion in the paper regarded severity of storms in the tropics. For this, he drew on a comparison of paleo-data, evidence of climate changes that caused the shifts between glacial and interglacial periods, with model predictions. Granted the "forcing" (climate scientists' word for what is causing stuff to happen) was different: a change in the position of the earth relative to the sun leading to a temporary imbalance in energy, rather than the greenhouse gases that do the forcing today. But if the same things happened to water circulation, carbon dioxide levels and global temperature, it seemed like a good idea to see what else was caused by these sorts of changes.

It turns out that there were major increases in storm severity at the glacial/interglacial boundaries. Back when there was a 6-9 meter sea level rise, as assessed by coral backstepping toward shore (coral only form in certain depths of water) and numerous other proxies for sea level height, there were also storms capable of whipping mega-boulders around in a way that calm ocean dynamics don't. The presence of these megaboulders on top of totally different rock formations is one piece of paleo-data. Cementation patterns, forming of hard rock out of sediments when ions are washed in, seen as shoals, ridges, and dunes, are also evidence of intense storm activity. Explanation: increased temperature differences with ocean depth cause increased eddy kinetic energy and also increase in a climate system property called baroclinicity. That means mismatch of air pressure gradients and temperature gradients. All of this is correlated with greater severity of storms.

And just like in Hansen's ideas of the present, at the glacial/interglacial transition, more meltwater led to reduced Southern Ocean ventilation. Here, not just heat ventilation but also carbon dioxide ventilation was also important. Carbon dioxide stays in the atmosphere a very long time, so it can be considered the control knob on climate. By contrast, radiation of heat due to clouds, trapping of heat due to atmospheric water vapor, and reflecting of heat by surface ice and snow are all rapidly adjustable (on a climate scale of hundreds of years, not on the scale of tomorrow's weather). Change the level of carbon dioxide in the atmosphere and you change the amount of warming. Consider known delays in these processes, and you get a good explanation of ancient climate change.

Last big idea: the model can be used to see if all these effects on the climate are a linear response to increased meltwater (more melting leads to proportional increases in everything else) or an exponential effect (because of positive feedback, each increase in meltwater flow leads to a much more massive increase in the changes it causes that gets still more massive with time.) Exponential effects can be thought of as having a doubling time: things are twice as bad in ten years, four times as bad in twenty years, eight times as bad in thirty, if the doubling time is ten years. The data matched an exponential better than a linear response. But meltwater flow measurements of the quality needed for modeling haven't been taken for long enough to be definitive. Hansen recommends more measurements. But the last time he said disaster was really likely to happen, it did, so the world should probably take "really likely" seriously from this guy and not wait for the level of proof needed to convict a tobacco company in court.

Got all this? Here are Hansen's conclusions, the only thing most news outlets picked up on: "High fossil fuel emissions this century are predicted to yield (1) cooling of the Southern Ocean, especially in the Western Hemisphere; (2) slowing of the Southern Ocean overturning circulation, warming of the ice shelves, and growing ice sheet mass loss; (3) slowdown and eventual shutdown of the Atlantic overturning circulation with cooling of the North Atlantic region; (4) increasingly powerful storms; and (5) non- linearly growing sea level rise, reaching several meters over a timescale of 50–150 years. These predictions, especially the cooling in the Southern Ocean and North Atlantic with markedly reduced warming or even cooling in Europe, differ fundamentally from existing climate change assessments."

If you understood the paragraph I quoted from the paper, I did my job. Now go yell at some politicians.

Recent Posts

See All

Terrorism and the Fortress Mentality

The event: September 11, 2001, al-Qaeda terrorists fly two planes into the World Trade Center, killing nearly 3000 people. The context: In 1991, the US tricked Saddam Hussein into invading Kuwait. As

Subscribe to this blog to get email notifications of new posts

Thanks for subscribing!

bottom of page